1 of 59

Natural Hazard Early Warning System Assessment

Table of Content


1 Preamble ............................................................................................................................. 2
2 Introduction ........................................................................................................................ 3
3 Responsible Organization in Georgia ................................................................................ 4
4 International Good Practice of FEWS .............................................................................. 10
5 Findings and Recommendations in Georgia .................................................................... 25
5.1 Available Knowledge Basis (Key Element 1 of FEWS) .......................................... 25
5.2 Flood prone areas of highest degree in Georgia ....................................................... 29
5.3 Monitoring and Warning Service (Key Element 2) ................................................. 35
5.3.1 Organization of the Hydrometeorological Department .................................... 35
5.3.2 Monitoring Service ........................................................................................... 36
5.3.3 Data collection .................................................................................................. 39
5.3.4 Field works ....................................................................................................... 40
5.3.5 Prediction Service ............................................................................................ 42
5.4 Dissemination and Communication (Key Element 3) .............................................. 46
5.5 Response Capability (Key Element 4) ..................................................................... 50
6 Conclusions ...................................................................................................................... 53
7 Appendix .......................................................................................................................... 59

2 of 59


1 Preamble

Based on the request of the Ministry of Environment Protection & Natural
Resources (MoE) of Georgia, UNDP assists the MoE in conducting a needs
assessment and a feasibility study for establishing an early warning system
for natural hazards in Georgia.

During the last decade, advances in science and technology have improved
the potential of early warning systems to reduce human loss from
meteorological or hydrological as well as from geotechnical disasters.

Effective early warning systems are widely recognized as good practice if
evidence-based and people-centered, and thereby empowering individuals
and communities to act on risks in sufficient time and in an appropriate
manner to reduce the possibility of injury, loss of life and damage to property
or the environment.

Putting early warning high on the national agenda is a strategic activity for
the organization at a time when climate change is likely to produce more
extreme climate events. In addition, this feasibility study is in line with the
second priority action of the Hyogo Framework for Action 2005-2015, that is
to identify, assess and monitor disaster risks and enhance early warning.

The current study focused on the flood warning subject but the prospect of
the integrated early warning system was also kept in view throughout the
project. The study was implemented in close collaboration with the UNDP
Disaster Risk Reduction Advisor, the UNDP Environment & Energy portfolio,
MoE, National Environmental Agency and other concerned national
authorities. The study includes the following main points

(i) Investigation of the current status of the flood warning
practice and risk analysis of other natural disasters
(ii) Makes recommendations for establishing early warning
system for flood hazards in Georgia and for improving
activities, which concern risk analysis of other natural
disasters
(iii) Recommendations on capacity development initiatives for
MoE and other key institutions for establishing and
operating an early warning system for natural hazards,
especially for floods
(iv) Discussion of the draft report with main findings and
recommendations of the study to key stakeholders
(v) Finalization of the feasibility study report with findings and
recommendations for various categories of stakeholders.

3 of 59

2 Introduction

On a global basis, floods have the greatest damage potential of all natural
disasters worldwide and affect the greatest number of people. The human
losses and damages caused by floods are also significant in Georgia. The
Appendix 1 shows statistics of natural disasters where one van see that the
flood damages have been exceeded 200 millions USD for the period 1995-
2008. The cases when the alert or flood levels were exceeded on the main
river gauges for the years 1986-2006 are presented in Appendix 2. The alert
water levels were exceeded more than 100 times while the flood levels more
than 50 times. We also need to note that the majority of the deaths
associated with flood disasters can be usually found within the most
vulnerable members of society, namely women and children. The greatest
tragedy is that most of these deaths can be either avoided or the number of
fatality fall can be dramatically reduced through investments in
infrastructures, preparedness activities, flood plain policy development,
effective watershed land use planning, upgraded flood forecasting and
warning systems and response mechanisms.

Therefore Georgian Government must undertake further measures to adopt
advanced approaches for flood disaster management. In order to promote the
flood security, the current Study focuses on the flood forecasting and
warning system and response mechanism, i.e. Flood Early Warning System
(hereinafter FEWS). The reason for the focus of this study is that flood can
be forecasted in real-time mode while for the other natural disasters can be
only given risk level but not quantitative values nor occurrence time.

The structure of the current report based on the key elements in the
Checklist of the Early Warning Systems published in Proceedings of the
Third International Conference on Early Warning (Bonn, 2006).

A complete and effective early warning system comprises four inter related
elements, which are as follows

1) Risk Knowledge
2) Monitoring and Warning Service
3) Dissemination and Communication
4) Response Capability

The Risk Knowledge based on the presumptions that the experts have
precise information on the locations and character of the natural hazards as
well as vulnerability of people (first priority) and objects of value or having
potential for secondary threats (for example flooded chemical storage) for the
people and/or environment.

The risks should be documented and these documentations should be easily
accessible. They ways and means for assessing the risks and the use of this
information will be discussed later because they are specific for different
4 of 59

natural disaster phenomena. We only note here that the risk information is
used for two purposes

To provide information on the location and
characteristics of natural hazard occurring in the
country in order to develop the protecting
measures (both of structural and non-structural
character)
To prepare operative rescue plans and to take
operative protection actions when the disaster
approaching or has happened.

The Monitoring and Warning Service lie at the core of the early warning
system. It is natural, that the earlier we know about the approaching
disaster the better we can protect the people and reduce the damages.
Therefore, the prediction tools are invaluable components of the early
warning systems. However, we have to also mention that there is no effective
prediction without well-functioning monitoring network. Therefore the
current paper will focus on both monitoring and predictive issues of the
Georgian Early Warning System.

The Dissemination and Communication is an organic component of FEWS
Services because the warnings should be delivered to authorities without
any delay when extreme flow is expected. However, the information
preparation is a complicated task because predicted primary data is not
informative enough either for authorities or public. Therefore the Warning
Service should make interpretation of the forecasted data and give an
evaluation of the expected situation with content and in form, which ensure
correct understanding. In case of emergency situation, the regular (daily)
information dissemination is not enough therefore special channel and
means of immediate communication should be planned and established. The
information to different users should be specific, for example operators of
hydropower plants request other knowledge than the rescue centers or
municipalities.

The Response Capability according the Checklist is the understanding and
respect of the risk warnings by the communities. However, the Consultant
Team looked also into an other aspect of this issue, namely the central
responsibility and task sharing between governing bodies such as Security
Council, Ministry of Environmental Protection and Natural Resources,
Ministry of Internal Affairs and Ministry of Territorial Development.

3 Responsible Organization in Georgia

Organizational and institutional arrangement for natural and man-made
disaster risk reduction and response in Georgia is regulated by the following
laws and legislative acts:

5 of 59

1. Law of Georgia on Emergency Situations - 17.10.1997 and its
amendment #2169-II of 05.07.2003
2. Martial Law of Georgia - 31.10.1997
3. Law of Georgia on Protection of Population and Territories Against
Natural and Manmade Emergency Situations - 08.06.2007 and its
amendment of 31.07.2009
4. Presidential Decree (#415) on National Response Plan for Natural and
Manmade Emergency Situations - 26.08.2008
5. Presidential Decree (#1061) on the Reorganization of the Legal Entities
of Public Law under the Ministry of Environment Protection and
Natural Resources of Georgia - 17.12.2005
6. Bylaw of the National Environmental Agency - as a direct follow up of
the above mentioned Presidential Decree
7. Decree #611 of the Ministry of Environment Protection and Natural
Resources of Georgia on the National Environmental Agency -
29.08.2008


These laws define responsibilities of a major national governmental
institutions (central, regional, or local) involved in all key activities that for
the moment are addressing hazard risk assessment, monitoring, information
dissemination and response actions practicable in Georgia.

The National Response Plan for Natural and manmade Emergency Situations
is a very comprehensive document that describes in details tasks of each
involved agency first of all in case of national level emergency situation, but
also responsibilities of regional and local operation teams. The plan defines
ways of overall coordination, relevant measures for disaster prevention and
mitigation, evacuation strategy, victims assistance and relief, liquidation of
consequences, etc. In general it defines 17 basic specific functions for
potential national-level emergency situation and assigns major and auxiliary
roles to key government agencies for addressing those functions. The Plan
also defines role to the Governmental Commission for Emergency Situation
Management, which in case of national-level emergency will act as a
supreme coordination body. At the same time executive role for a crisis
prevention is mandated to the Interagency Crisis Prevention Operation
Center, which will operate under the Emergency Management Department
(EMD) of the Ministry of Internal Affairs.

As stated above however, this scenario is mostly applicable in case of
national-scale emergency situations taken that local and regional authorities
are not capable to handle emergency situation themselves, or their
competencies are beyond of existing level of disaster (devastating
earthquakes, epidemic outbreaks, volcanic eruptions, etc.). In all other cases
responsible authorities, both for coordination, as well as for crisis prevention
and field operations are sole responsibilities of the regional (or autonomous
republics of Adjara and Abkhazia), and local self government authorities.

6 of 59

According to the classical people-centered Early Warning System approach
for disaster risk reduction, adopted by the World Conference on Disaster
Reduction (18-22 January, 2005), also known as Hyogo Framework, the
following four major key elements are applicable to Georgia as well: a) Risk
Knowledge; b) Monitoring and Warning Service; c) Dissemination and
Communication; d) Response Capability. Thus, our attempt to assess
institutional set-up of the Early Warning System in Georgia is based on the
analysis of existing system in accordance to the above mentioned scheme.

It should also be noted that since consultants where limited in time and
therefore forced to choose more focused theme for their investigation,
disasters caused by hydro-meteorological processes have been selected as
main subject of the assessment. Consequently, from a large number of
related to the disaster management stakeholders' list (attached) the following
important institutions have been targeted as key players for risk knowledge
assessment/monitoring and immediate response: the National Security
Council of Georgia; Emergency Management Department of the Ministry
Internal Affairs; Ministry of Regional Development and Infrastructure;
Ministry of Environmental Protection and Natural Resources; National
Environmental Agency.

Office of the National Security Council of Georgia

The National Security Council is an advisory body directly subordinated to
the President of Georgia. As it is stated in the Article 5, Paragraph 2 of the
Law of Georgia on the Emergency Situations (17.10.1997), as well as Martial
Law of Georgia (31.10.1997), the Security Council is a main coordination
body in case of national-level emergency situation for any kind of disaster
prevention and response activities.

The Security Council currently coordinates activities to investigate legislative
needs for crisis management in Georgia, as well as so called 'Disaster
Mapping', or the risk knowledge assessment study - identifying natural
hazards' sensitive areas, communities vulnerable to the hazards, type and
character of the risks. First version of the Disaster Mapping study is
expected to be available by early next year.

Although Political 'weight' and influence of the Security Council is quite
significant, it lacks institutional and human resources for actual
implementation of assigned to it functions. Thus their capacity as disaster
risk reduction coordination body currently is rather limited. It is
recommended therefore that before modernized and institutionally
strengthened the Security Council should play more supervising and
controlling role in disaster management than executive, or administrative.

Ministry of Environmental Protection and Natural Resources of Georgia
(MEPNR)

7 of 59

MEPNR in this scheme is represented by the National Environmental Agency
(NEA) - a legal entity of public law operating under the Ministry of
Environment and the Department of Sustainable Development, Division of
International Relations of the Ministry itself. Structure and functions of NEA
will be described in details in the chapters bellow (Chapter 5 and beyond),
therefore here only state that in the classical people-centered Early Warning
scheme the role of NEA clearly corresponds to Key Element-2: Monitoring
and Warning Service. However the consultants believe that their capacity as
contributors for Key Element-3: Dissemination and Communication, i.e.
delivering warning messages to proper audience in a timely manner, should
be increased in a future. This will becomes more obvious if considering the
fact that the predecessor of the NEA's Hydromet Department - the National
Hydro-meteorology Service that had much advance technical capacity and
extensive observation network, was also responsible for dissemination of
information on hydro-meteorological hazards during the former Soviet
system.

On the other hand, role of the Department of Sustainable Development
(DSD) and its International Relations Division, being Focal Point for the
Hyogo Protocol, is not clearly defined. Having vast experience in reporting
and assessment processes for different international secretariats and
conventions we would think DSD as major contributor for Key Element 1:
Risk Knowledge study, as it has closer and more frequent interaction with
the data producers and analytical think-tanks. In this scenario we would
leave for the Security Council role of overall supervisor and coordination to
link four different Key Elements to each other.

Ministry of Regional Development and Infrastructure (MRDI)

MRDI was established in January 2009 by special Governmental Decree. It
was created on the base of the existing State Minister's office for Regional
Development, as well as Infrastructure and Transport Administration
Departments of the Ministry of Economic Development. Recently, as the
country's priority shifted towards decentralization and strengthening of a
regional self governance, the MRDI's role relative to other ministries for
disaster risk reduction and prevention has significantly increased. According
to the amendment N1566 of 31,07.2009 to the Law of Georgia on Protection
of Population and Territories Against Natural and Manmade Emergency
Situations, MRDI become responsible for any disaster risk reduction and
prevention activities on regional and local levels. Namely, role of a Governor -
State Plenipotentiary increased to the following responsibilities:

- issuing individual legal acts to ensure protection of population
and territory against emergency situation;
- coordinating activities of relevant institutions for prevention of
consequences caused by emergency situation at relevant
administrative-territorial unit;
- providing use of regional emergency situation response forces in
accordance to the National Response Plan of Georgia;
8 of 59

- coordinating evacuation of population in case of emergency
situation;
- coordinating activities to provide temporary, or stationary
shelter to evacuated population;
- receiving and processing relevant information for protection local
population and territories against emergency situation;
- coordinating and supervises distribution of humanitarian aid;
- coordinating rescue and other urgent operations;
- coordinating secure functioning of vital regional institutions
during emergency situation;
- providing other relevant activities to prevent emergency situation
if possible and respond to the consequences in accordance to
the Georgian legislation.

Since April 2009 the MRDI is working on a new legal document according of
which the Ministry will establish Regional Service Centers of Disaster
Management in each region of Georgia. The Ministry will hire and equip
about 30 professional staff per Service Center that will be responsible for
and unify three functional units: 1) data collection and processing, 2) risk
assessment and contingency planning (about 5 staff members in each), and
3) disaster response and rescue operations (21 staff - 7 rescuers per team of
three shifts). The MRDI plans to provide very intensive training with the help
of the Ministry of Internal Affair's Emergency Management Department for
new recruiters. In addition, after establishment early next year, the Service
Centers will be equipped with the latest available technology and necessary
data and methodology tools.

Thus, as it seems from the abovementioned, once MRDI's Regional Service
Centers will gain full capacity they should be able to cover at least two out of
four Key Elements of the Early Warning System scheme: Dissemination and
Communication of the early warning information (Key Element 3) and have
Response Capability (Key Element 4) on place to prevent/mitigate disasters,
increase community preparedness, strengthen their response capacity,
increase public awareness and liquidate consequences of natural or
manmade hazards on regional and local levels. However, it should be also
noted that this will be very costly and time consuming goal to achieve.
Therefore, the government should think about alternative ways to fill the gap
meanwhile.

Emergency Management Department (EMD) of the Ministry of Internal
Affairs

From the above mentioned institutions involved in the early warning and
natural hazard assessment and response network in Georgia EMD is
probably the most organized, best equipped a trained institution, having very
long experience of handling emergency situations caused both by natural
disasters, or manmade catastrophes. Their role as a major Early Warning
System player is particularly significant for information dissemination &
communication activities (Key Element 3), as well as for response capability
9 of 59

(Key Element 4). Therefore EMD is listed as a key coordination and
implementation body for the most functions of the National Emergency
Management Plan.

EMD has excellent training infrastructure and analytical capacity, as well as
perfectly arranged communication network with the data collection and
knowledge base assessing agencies. In case of national-level emergency
situation EMD has legislative mandate to act both as an Interagency
Operation Center for the Crisis Management, as well as the Field Operation
Center located directly in the crisis area, or in its vicinity. Although their role
during the regional and/or local scale disasters is limited by rescue
operations, the consultants believe that Georgian government should use
benefit of having EMD as center of excellence and utilize their resources
more widely before MRDI's Disaster Management Regional Centers will gain
institutional strength and all necessary skills to address the mentioned
issues.

Overall the national response plan can be schematically presented as shown
on the Figure 1. bellow: with clearly defined hierarchy and links in case of
National, or Local/Regional scale disasters, but at the same time very week
inter-linkages and sometimes none existing relations between the agencies
from the top (the National Security Council) to the bottom (regional
authorities, local municipalities).



Figure 1. Scheme of Institutional Arrangement for Emergency Situation
Management on the National and Regional Levels
10 of 59


4 International Good Practice of FEWS

The structure of the presentation of international good practice will follow
the Checklist presented above.

Knowledge Base (Key Element 1)

As it is mentioned above the risk knowledge points at the locations of
potential natural hazards, describes their character, extensions, frequency
and group of people and objects can be hit. The risk documentation should
support inter alia operative protection actions for case when the disaster
approaching or has happened. The knowledge of the potential hazards
makes possible to be prepared for natural hazards by

delineating the exclusion zones, for example, where
no houses are allowed to be erected
building protection structures
focusing on EWS development for the regions
vulnerable for specific natural hazards
preparing emergency plans for settlements,
infrastructure industry plants and agriculture units
storing rescue equipment and other means of relief in
the vicinity of potential natural hazard locations
training the organizations and public to take
measures for minimizing the risks and to correct
behavior when the disaster is a factum.

The risk maps and other records are very important decision-support means
when the disaster is approaching or is a factum. These documents allows the
emergency centers

to forward direct information to public, to infrastructure
operators, to industrial plants etc. known from risk maps
or other records that they are under risk
to plan operative evacuations based on the knowledge
which roads, bridges, tunnels are still useable at certain
flood levels
to plan and take protection measures on objects being at
risk.

The first step for implementing FEWS is the system set-up. During this
process, the river basin should be systematically analyzed and matters as
below precisely defined

- water regime of river system with special attention to the high flows
- existing observation network
- available historical data: hydrological and meteorological
11 of 59

- hydro-morphology
- maps incl. GIS formats
- hydro-technical structures influencing the river flow
- communication facilities
- development plans concerning settlements, infrastructure,
hydrotechnical constructions, communication facilities
- objects at risks such as settlements, industrial plants, agricultural
fields and farms, infrastructures and any other object having potential
for secondary dangers (chemical stores, fuel tanks etc.).

The analysis of above mentioned factors helps the R&D personnel to define
the needs for the FEWS, the possibilities to set up and operate it.

Concerning the knowledge base of floods, we have to note that Directive
2007/60/EC on the assessment and management of flood risks entered into
force on 26 November 2007. The Directive requires Member States

(i) to assess all water courses and coast lines at risk from
flooding,
(ii) to map the flood extent and assets and humans at risk in
these areas and
(iii) to take adequate and coordinated measures to
reduce the flood risk.

The main aim is to reduce and manage the risks that floods pose to human
health, the environment, cultural heritage and economic activity. The
Directive requires Member States to first carry out a preliminary assessment
by 2011 to identify the river basins and associated coastal areas at risk of
flooding. For such zones they would then need to draw up flood risk maps
by 2013 and establish flood risk management plans focused on prevention,
protection and preparedness by 2015.

A handbook on good practices for flood mapping in Europe was prepared by
EXCIMAP (a European exchange circle on flood mapping) Endorsed by
Water Directors, 29-30 November 2007. The Handbook contains
descriptions of the different types of flood maps therefore there is no need to
give all details here. What we can mention here is that hazard maps will give
the physical characteristics of the flood while the risk maps define the group
of people, settlements, industrial objects, infrastructure and agricultural
land areas, which can be potentially hit by the flood.

In response to the Flood Directive the EU member states have started
complex works on the flood plains to delineate the areas which are at risk.
The tasks include processing of data hydrological, producing digital terrain
models (DTMs) and performing hydraulic computations. Figure 2 presents
an example inundation map, which was calculated at different water levels
along River Tevere (Italy). Red shading shows the deepest areas, which are
most frequently inundated. Yellow shading shows those parts of the flood
plain, which are inundated by flood of middle size. While the green areas get
12 of 59

under water only in case of extreme high floods occur. The Figure 3 shows a
map created in North of Sweden where the objects in the risk zone are
identified (insert).

We should note here that the preparation of flood maps requests advanced
hydraulic computations, which in turn use extended databases containing
of data for hydrology, hydraulics, hydro-morphology, hydrotechnical
structures and relief (DTM). Well-known hydraulic software tools for flood
mapping are, for example, MIKE 11 or HEC RAS. These programmes are
capable to simulate the flow in the river bed and on flood plain. Their GIS-
based data management systems handle the input and outputs into/from
the computation modules.

In the end, we give an outline of suitable scales of flood maps for different
user purposes: (i) Scale of the order of 1:500,000 to 1:1.100,000 is feasible
only for large flood plains and gives overview information on village and town
level, only the approximate population per municipalities, villages or towns
can be represented. (ii) Medium scales (1:100,000 to 1:250,000) can be used
for broad-scale infrastructure like road or rail network, or agriculture and
forestry. (iii) Scales of 1:5,000 to 1:25,000 are for city or village plans,
detailed information about individual buildings, social structures or social
groups or about individual facilities require maps with a large scale.



Figure 2. Inundation map on the flood plain of River Tevere (Italy).
Shading of inundated areas: red frequently; yellow at average high flow;
green at extreme high flow.
13 of 59





Figure 3. Flood risk map along the shore line of Lake Stors (Sweden). Red
shading shows inundation at low level water stage while higher water can
reach the objects shown and tabled.

Monitoring and Prediction (Key Element 2)

Monitoring

Monitoring and prediction service is the core of the FEWS and consists of
two organic components, i.e. the observation network and the prediction
bureau.

Diverse inputs into the FEWS are shown in Figure such as ground
measurements from hydromet network, signals from radar locators, remote
sensing information and forecasts from regional meteorological weather
systems.

General modules prediction tool and their input needs are shown in Figure
and a complete monitoring and prediction system is presented in Figure The
main tools of the observations are rain and river gauges on the ground,
which networks can be completed by remote sensing and radar locators
14 of 59

(Figure ). The ground and remote sensing data forwarded to the, the Water
and Power Development Authority (WAPDA) and Provincial Directorates (PDs)
as well as to the National Flood Forecasting Bureau (NFFB). One can see
that some telemetric stations are directly connected to the SAMO and
SOBEK forecasting model units run by the NFFB.

The design of FEWS starts with collection of historical hydromet data which
information subsequently should be processed and augmented where it is
necessary. The operational rules of the hydro-technical structures working
in the river basin should be also obtained from the operators of these
objects. The data analysis will also point out the further needs for upgrading
of the monitoring network: more gauges to be deployed on the ground,
remote sensing data to be included, telemetric stations to be installed etc.

The effective data handling within the forecasting center can be ensured by a
well-designed and safely functioning data management system. In order to
issue the warnings as early as possible, the data should be transferred in
online mode to the forecasting center where the processing should be also
rapidly made and introduced into the forecasting models.

Today, the FEWS packages include data management platforms, which
contains of interfaces for receiving the data from weather forecast systems,
of control of automatic telemetric instruments and of handling central
databases in the hydromet centers.

15 of 59



Figure 4. Information flow to the Hydrological Forecast Service.

Prediction

There are many different flood warning and prediction systems set up in the
World. The main features of a such system depend on a series of factors,
inter alia as

- physico-geographical condition of the region
- water regime in the river basin
- status of the monitoring network
- equipment base
- software tools
- communication facilities
- competence of the staff taking care of the monitoring, field
works, data collection/processing and forecasting
- available finances etc.

The phyco-geographical properties of the river basin and the end-users
needs in early warning will tell to the future developers and operators which
kind of models she/he should include into the FEWS.

16 of 59

Economical or other constrains may challenge the future owner or operator
of the system to exclude one or more elements or reduce their capability.
Such attempts result always errors in FEWS function.

We will shortly review the main features of flood warning and predicting
systems. However, we have to emphasize clearly that this paper not going to
propose any system for implementing in Georgia. There are many reasons for
this stand point inter alia the introduction of software and hardware systems
is accompanied by commercial interests, which aspect should be carefully
avoided in this paper. In addition, the selection of early warning system as
well as its components (observation and measurement equipments,
communication facilities, data processing and storing techniques, software
tools etc.) should be made in a carefully controlled procurement procedure,
which fulfils all relevant legal requirements.


Modules of the prediction unit are selected based on the geographical
location of the area we need to supply the forecasts for. If the region we are
dealing with is the source area of a water course then first component in the
model system is the rainfall-runoff calculator. The request of such calculator
is a network of densely deployed rain gauges with frequent and online
transmitted measurements of rainfall intensity to the FEWS center. If such
rain gauge network is not available then there is no any possibility to deploy
and operate such FEWS, which includes rainfall-runoff calculations because
there is no any chance to make numerical assessment of the runoff based on
rainfall quantities. There are some other options for qualitative (approximate)
estimation of extreme flow, which may be caused by rainstorms. These
possibilities we will mention in this paper later.

The next step is calculation of the propagation of flood wave in the river bed
from upstream reach to the middle and lowland river stretches. Thus, the
rainfall already run on the soil surface and reached the river bed in the
mountainous part of the watershed. In turn, the high flow is moving
downstream. There are so called physically based models to calculate how
the flood wave moves along the river. All these models include the same
hydraulic equations. The differences between these models are how the user
(i) makes data input, (ii) controls the computational runs and (iii) receives
the results (tables, graphics and maps). All these models can provide lot of
information in their outputs: the size of the flood wave, time when the peak
flow will arrive to different locations along the river, maps of inundations
areas on the floodplain etc.

However, the future operator of the FEWS should consider that physically
based models request a lot of hydrological, hydro-morphological, hydraulic
parameters for proper functioning. If such database is not available there is
not any sense to waste the time on the implementation of such models
because they will never function correctly. In case of lacking data for
physically based models, the operator should use simplified methods (so
called flood routing) or statistical relations between the upstream and
17 of 59

downstream river gauges. However, it does not mean that the FEWS can
function without adequate hydro-meteorological data even in case simplified
models are included in the system.

Additional tasks are to be solved in the river basins where hydro-technical
structures influence the flow regime (reservoirs, flood gates, spillways, flood
channels, flood detention basins etc.). In this case, the FEWS operator
should take into account the effects of these structures on the flow regime
and the operational regime should be incorporated in the FEWS. We have to
emphasize the importance of components accounting for effects of
hydrotechnical structures because these objects may strongly affect the flow
regime and they have special management schemes in case of extreme high
flow.

When flow regime studied, the available information assessed, the needs are
clear then the FEWS installation procedure can proceed to the next phase,
i.e. to selecting the models to be used and coupling the components to each-
other.



Figure 5. Flow chart of the early warning process forecasting phase.

The FEWS predictor consists of three major parts (Figure ):

The Calibration System of the Model will determine the parameter values.
The procedure uses historical hydro-meteorological data measured and
observed in the river basin. This work should be done when the FEWS is set
up and then should be repeated regularly (updating).


Rainfall
measurement
and forecast
Flow regime
calculation
along the river
Rainfall-
runoff
calculation
Hydro-
technical
structures
modify the
flow
Flow data,
calculation
along the
river

Flow data along
the river
Flow data along
the river
State of the
hydro-
technical
structures and
operational
rules
Rain gauges
and forecasts
18 of 59



Figure 6. Complete forecasting system.

19 of 59



Figure 7. Prediction module.

The Operational Forecast System (OFS) should solve two tasks, namely,
providing information on the current status of the watershed and forecast
future conditions. Hydromet information introduced into OFS are
meteorological and hydrological data observed on the ground as well as
forecasted meteorological values taken from regional prediction systems. All
hydromet data before introducing it into the operational forecast models
should be checked and processed. The chain of meteorological-hydrological-
hydraulic models should be online connected, however, interactive regime,
i.e. option for specialists to adjust the computation process should be given.

Finale stage of the prediction phase is the evaluation of the forecasts done in
Extended Stream Flow Prediction System. One should keep in her/his mind
that the forecasted values include a grade of uncertainty because of a long
range of reasons, inter alia, limited hydromet information available, errors in
the obesrvations and necessary simplifications in the used methods.
Therefore, each time we hand out hydromet information we have to
accompany with so called confidence interval, i.e. a gap within that the real
water level (discharge) can occur.

Instead of giving here a long list of available hydrological forecasting models
(which anyway easily available on internet) we focus on two advanced models
and on a hydrometeorological data integrating platform, which were tested in
Georgia.

20 of 59


Operational Flood Early Warning Systems Tested in Georgia

MIKE 11 is a software package provided by Danish Hydraulic Institute, DHI
(http://www.dhigroup.com/Software/WaterResources/MIKE11.aspx ). The
hydrodynamic module (HD) is the core of MIKE 11. It provides a library of
methods for computational simulations of extreme flow on flood plains. The
package includes NAM hydrological model for rainfall-runoff computations,
hydraulic model for flow transformation on the river channel and water
movement of the flood plain. The later one allows the user to create
inundation maps.

In the frame of a project supported by Science and Technology Center in
Ukraine (STCU) the MIKE 11 package along with computers were
transferred to the Faculty of Natural Sciences, Tbilisi State University. The
MIKE 11 was tested on a reach of Rioni River in Alpana region, where
inundation maps were created for flood flow of different probability (Figure ).
A new small research project is in process right now for testing the model on
the Natanebi River in Guria Region. The hydro-morphological measurements
of the river bed and mapping of the flood plain is in process.


Figure 8. Rioni River in Alpana Region.
Inundated areas assessed by MIKE 11 model
(probability of the floods: 1%, 5%, 10% and 20%).
21 of 59


As it is mentioned above, the hydraulic computations need extended hydro-
morphological database, which generally are not available in Georgia.
Therefore, the MIKE 11 hardly applicable for operative forecasting purposes
but it can be used for flood mapping and consequence analysis of extreme
floods including those induced by failure of hydrotechnical structures.

The Delft Hydraulic Institute provided its Delft-FEWS package has been
used in a series of projects world-wide. The Georgian National
Environmental Agency (NEA) is testing the Delft-FEWS package therefore we
give a brief review of it while the status of the testing process given in section
5.3.5.

The full Delft-FEWS software package consists components as

(i) Interfaces connecting to external data sources
(ii) Validating, interpolating and transforming data
(iii) Simplified forecasting models
(iv) Advanced forecasting models and tools
(v) Presentation and dissemination of the forecasts

Delft-FEWS provides an import module that allows to import of on-line
meteorological and hydrological data from external databases. These data
include, for example, time series obtained from telemetry systems such as
observed water levels, observed precipitation, but also meteorological
forecast data, radar data, etc. The import of external data also supports
ensemble weather predictions, such as those provided by the European
Centre for Medium Range Weather Forecasts (ECMWF).

Delft-FEWS performs some quality checking of data and provides support in
transforming data with disparate spatial and temporal scales. This includes
geo-statistical spatial interpolation to derive for example areal weighted
precipitation from spatially distributed point sources, or from spatial data
such as radar data and numerical weather prediction models.

The simple forecasting modules are based for example on relations between
upstream and downstream gauges while the more advanced models are also
available. The results are presented in tables, graphics and maps. The
testing exercise at NEA was made using the HEC-HMS model as predictor,
which system was set on the Delft-FEWS database manager.


22 of 59



Figure 9. Inundation map assessed by HEC-RAS model for different water
stages in the river.

Model HEC developed by U.S. Army Corps of Engineers, is a complete Flood
Hydrograph Package, which allows the user flexibly compose the software
tool package of modules for the task. The HEC-HMS (Hydrologic Modeling
System) is a real-time flood forecasting version, which simulates the rainfall-
runoff process in a catchment, transforms the flood wave along the river and
accounts for the impacts of hydrotechnical construtions.

Different components of HEC-HMS are available in the tool box, which may
also simulate the effects of dam breaks and inundation maps (Figure 9.). The
HECHMS allows the choice of different formulations of the hydrologic
processes involved (Figure 10.).

(i) surface runoff response to precipitation
(ii) snowmelt: temperature-index method
(iii) transformation of flow in complex river channels.

23 of 59



Figure 10. Interactive composition of the river network by HEC-1f.

The great advantage of the HEC-HMS is that the user can freely compose the
modules, which represent the river elements (junctions, diversions etc.),
hydrotechnical objects (reservoirs, detention basins etc.), which are located
in the river catchment.

HEC-HMS is applicable in a wide range of geographic areas for solving the
widest possible range of problems. This includes large river basin water
supply and flood hydrology, and small urban or natural watershed runoff.
Hydrographs produced by the program are used directly or in conjunction
with other software for studies of water availability, urban drainage, flow
forecasting, future urbanization impact, reservoir spillway design, flood
damage reduction, floodplain regulation, and systems operation.

In summary, first we have to mention concerning the above mentioned
world-wide used software tools that the MIKE 11 is a commercial product
while Delft-FEWS and HEC freely available, however, throughout training
24 of 59

and qualified supervision are required when introducing MIKE 11, HEC and
Delft-FEWS.

A major point is that the effective use of these and other models requires
well-skilled personnel. The operators should be well-familiar with
hydrological and hydraulic processes, should be fully computer literal,
she/he must be experienced in hydrological/hydraulic modelling and in
application of GIS technique. Therefore, special long-term training is needed
for the group, which should be of interdisciplinary character.

Another major point is that long-term set-up, calibration and testing period
is needed before any of these tools can be introduced in the operational
practice. Therefore, well-planned personnel policy is a basic assumption for
operational implementation of these modern software tools.

Data dissemination (Key Element 3)

It is exceptionally important that the information of extreme events reaches
the population being at risk, authorities to take all necessary measures to
protect the people and their properties, hinder industrial damages, which
can lead to severe environmental catastrophes. The data dissemination
theme made of two components, namely, the technical facilities and the
institutional structure set up for informing the population and other parties
involved in responding for extreme events.

The present paper described above the basic functions of a national
hydromet services, i.e. running the monitoring networks; collecting and
processing the observations, calculating the forecasts.

The next task that the warning should be forwarded to the governmental
organization and those take the necessary protection measures and inform
the relief organizations, local authorities, people, public utilities, industrial
plants, farming associations etc.

When the computation phase of the early warning activity is ready and the
Service Provider of Early Warning Service issues the alert signal, there a very
important tasks to take care of:

(i) professional analysis of the current status in the river
basin and its interpretation for the broader range of
stakeholders
(ii) the analysis of the forthcoming situation and its
interpretation for the broader range of stakeholders

The Figure 11 presents schematically the characteristic river states: normal
flow there is no need for any special actions; monitor (alert) stage when the
water level is approaching critical level more frequent observations should
be made and some additional water gauges should be checked; flood stage
when the river enters the flood plain emergency alert should be issued
25 of 59

based on the forecasts, i.e. before the water level reaches the flood stage. In
fact, Georgian hydromet practice exactly follows this pattern.



Figure 11 Water stages requiring different preparedness
from the FEWS and respond staff.

Response Capability (Key Element 4)

The Response Capability according the Checklist is the understanding and
respect of the risk warnings by the communities. It is a major point for the
effective FEWS that the information should be adjusted to the recipients
needs. For example, it is obvious that the public information must be
different from the data exchange between the Hydromet Service and
operators of the hydropower plants. Section 5.5 outlines the hierarchic
information exchange between different Georgian agencies responsible for
the FEWS incl. the relief operation.

5 Findings and Recommendations in Georgia
5.1 Available Knowledge Basis (Key Element 1 of FEWS)
The Consultant Team has visited institutes and made themselves acquainted
with available information concerning the knowledge base, R&D works and
other early warning related activities performed at Georgian Institute of
Water Management, Faculty of Natural Sciences of Tbilisi State University,
Institute of Geophysics, Seismic Monitoring Center of Georgia, Regional
Environmental Center for the Caucasus (REC) and National Environmrntal
Agency.

The studies of natural disasters have been done for decencies and performed
at high scientific level in Georgia. Therefore we can firmly state that there is
a good knowledge base available to tackle the natural disaster reduction in
Georgia. The other positive aspect in this subject is that the institutions
show definitive signs of recovery after the severe difficulties experienced
during the transition period. However, the shortage of funds is still obvious,
which result lack of personnel, limited access to modern equipments and
reduced monitoring programmes. Another notice is that the Research and
Development Activities (R&D) are not coordinated to the necessary extent,
which fact hinders fully utilizing the available knowledge of Georgian
experts. The reasons for latter obstacles can be several, inter alia, the funds
are too small for joining the projects in programmes, the funds come from
26 of 59

different sources (incl. aid, EU R&D programmes, national finances etc.) and
therefore it is difficult concentrate them.

The deepest study of existing situation was made at Hydrometeorogical
Department of the National Environmental Agency (NEA) being official
operational warning service for the natural disasters (except the
earthquakes) in Georgia. However, we have to clearly mention that R&D
activities performed at other organizations are also very important
contributions to the disaster reduction in Georgia.

National Environmental Agency

National Environmental Agency (NAE), is dealing with all natural hazards
(except the earthquake) and the leaders and staff make all efforts to save the
documents, observations, measurements have been collected for
decenniums. The other objective is that continue the measurements, data
processing and assessments to update the information on the current
conditions of hazardous sites. The new hazard inventories utilize the
advanced technologies, for example computers, digital data management,
GIS, telemetry, modeling etc. The Hydro-meteorological Department focuses
on atmospheric events (snow, rain, temperature, wind) and their extreme
consequences (flood, drought, windstorm, rainstorm, avalanches), while the
Geological Department focuses on mudflow, landslides and erosion. The
natural hazards are recorded in different catalogues and year books. The
information collected provided base for gross delineating of the hazard areas
and are available at the National Environmental Agency.

A Year Book issued annually, which includes information ad assessments
concerning the hazard conditions for mudflow, landslides and erosion. The
maps are available in scales: 1:500.000; 1:200:000; 1:50.000 and 1:25.000
for the entire territory of Georgia, while for 50% of the country: 1:10.000. In
addition, there maps of 1:2.000 (design solutions) available for around 100
landslide places.

Division of Hydro-meteorological Hazards and Damage Mitigation collected
and systemized important information: Catalogues of Natural Hazards,
which were completed for floods, droughts, hails, windstorms and
avalanches. The Catalogues were based on information from many different
sources such as hydromet observations, historical documents of disastrous
events, old newspapers etc. Based on the information were revealed the date,
location and extension of the disasters as well as the consequences (losses of
people and damages).

The information concerning the natural catastrophic events is of two types.
One of them lists those events, which the NEA experts were able to trace in
historical documents, news papers etc. The underlying reasons for disasters
and the caused losses and damages were reconstructed by applying direct
and indirect searching methods. The second sort of information is based on
27 of 59

regular hydromet observations and contains of recorded extreme high water
discharges, their staring and ending dates and their duration.


At the end, we have to mention that the Hydro-meteorological Department
has been dealing with an extremely important work, namely, systemizing
and saving invaluable hydromet measurements, time-series, and
observations in digital form on magnetic holders. The works has been
completed to 60%, however, it is still lot works should be done before is too
late because it is more and more difficult to save, handle and read the
archive papers.

Vulnerability Gradation of Regions in Georgia

The vulnerability gradation for floods within Georgia stems from the Regional
Hazard Mapping Programme coordinated by Nodia Institute of Geophysics
and supported by grants from EUR-OPA Major Disasters Agreement and
Georgian National Grant Foundation. Since, the Programme covered South
Caucasus and therefore institutions participated from Armenia, Azerbaijan
and Georgia. The Regional Hazard Mapping Programme made an great
attempt to avoid the problems encountered by Global Disaster Risk Maps
Programme prepared by IBR, WB and Columbia University. For example, the
Hotspot Map of Global Mapping considered South Caucasus as prone only to
hydro-hazards omitting geophysical ones. This misjudgment was caused by
the coarse scale of global mapping and insufficient information from where
Natural Hotspot Map was created, namely, GDP statistics and probability
analysis of hazard occurrence.

The Regional Hazard Mapping Programme collected information in the
administrative districts of Georgia and created a valuable database by
compiling data of economical statistic, losses caused by disasters etc. Lot of
assessment works were based on indirect information because the primary
data was scarce or not available.

The Regional Hazard Mapping Programme created small scale maps not only
for floods but also for earthquakes, landslides, mud and stone flows,
avalanches and flash floods. The gradation of these hazards was placed on
map 1: 1.000.000. As we mentioned in Section 4, maps of such scale provide
overview information but does not fit to risk assessment nor operational
disaster reduction.

The flood work was done within the Regional Hazard Mapping Programme
when hydro-meteorological data observed between 1961-2000 was
processed. The flood criteria for maps was that water level of annul
maximum flow with return period of 10/100 years equal or exceeds the alert
level at the given gauge. The alert water levels on the river gauges were
defined under the soviet time based analysis of the adjacent areas, i.e. if the
water level exceeded the alert level then living houses, industrial or military
objects and agricultural land would be at risk. The underlying technical
28 of 59

documents from Soviet time are not available anymore, however, there is no
better criteria today therefore still those alert levels are accepted as base for
marking the hazard zones on the floodplains.


Further flood mapping works were also discussed in details at the NEA
Hydrometeorological Department. Maps of scale 1.50.000 were digitized for a
part Svaneti, part of Mtskheta-Tianeti and Aragvi (white) and are in process
for upper and lower Svaneti and part of Adjara. The objective for
digitalization is a plan to make maps for all natural disasters, inter alia, for
floods. The preparation of flood hazard and risk maps is a very good
initiatives but this work suffers from some constrains. The major obstacles
encountered are

(i) too few experts are available at the NEA Departments
skilled for this type of works,

(ii) only paper maps are available for preparing Digital
Terrain Model (DTN) therefore the GIS-specialists is
manually digitizing maps of scale 1:50.000. Manual
processing work is very time consuming and his skill
could be used in a much effective way if the digital maps
were available,

(iii) the resolution of the map is too small. The scale
1:50.000 was chosen because even digitizing these maps
takes very long time. However, resolutions for flood maps
should be in vertical 20 cm and in horizontal around 10-
29 of 59

20 m, which requirements maps of scale 1:50.000 cannot
satisfy. In addition, the technology of preparing flood
hazard and risk maps is not clear at the moment for the
NEA staff.

5.2 Flood prone areas of highest degree in Georgia

The full reconstruction costs of the Georgian hydrometeorological service
were estimated by a World Bank study () and set to 10 million USDs. The
current study makes an attempt to find the minimum investment level,
which would eliminate those gaps, which first of all hinders the effective
function of the early warning system of natural disasters. We are focusing on
hydro-induced events or on those amplified by hydro-meteorological factors,
such as flood, mud flow, landslide and avalanches. The Consultant Team
aims at utilizing the already deployed hardware and software tools at the
NEA (first of all at Hydromet Department). Therefore significant part of
suggested measures are closely linked to available new equipments and
software tolls, which are not in full operation. In order, to set the priority
measures, first we look the vulnerable areas and the installations there.

Floods of large territorial extension in Georgia

The most sever and extended floods are caused by coincidence of intensive
snowmelts and heavy rainfalls. The only rainfall induced high flows are
observed during the period October-November (and sometimes December).
Rainstorms induce floods in Adjara during winter season as well as during
other seasons.

The spring flood is forecasted based on the snow resources in the
mountains, which will smelt away during a longer time. Consequently, the
spring flow can be also forecasted for a longer time ahead.

Mtkvari River Basin

The so called historical maximum flow occurred in 1968 in the Mtkvari River
Basin. The weather became warmer during the second decade of April when
intensive precipitation fell over the river basin where substantial snow
resources were mounted already. The snowmelt was intensive and as a
result the water levels exceeded the historical maximums at majority of the
river gauges functioning along the Mtkvari River and its tributary at that
time.

The losses were significant, inter alia Khashuri-Akhaltsikhe railway line was
flushed away in Borjomi, bridges and roads were damaged, living and
industrial buildings were broken, agricultural lands were inundated.

One of the most extreme large scale flood occurred in 2005 in the Mtkvari
River Basin when during the winter substantial snow resources were
30 of 59

deposited in the mountains. The snow recourse was as large as of 4%
probability, average depth reached 2.70 cm. The largest snow resources were
in the Aragvi River Basin. In addition, intensive rainfall have covered the
region during the days 25-27 April, 2005.


Figure 12. Aragvi River Basin with Zhinvali Water Reservoir.

There was a prediction for 3-4 days ahead that warmer humid air masses
approached the region therefore flood warning was issued. Zhinvali Water
Reservoir (Hydro Power Plant) released water and made a large volume free
for regulating the incoming flow. Thanks to this regulation, the peak flow at
Tbilisi was 2.200 m3/s (565 cm) and stayed below the historical maximum
2450 m3/s (722), which was observed in April 1968.

31 of 59

Great amount of rain fell down over the Mtkvari River Basin on 4-7 June,
2005. The basin average was 68 mm while on some places exceeded
100 mm. As a result, water level in Mtkvari River rapidly rose and exceeded
500 cm at Tbilisi, which corresponded 2250 m3/s. Based on the short-term
hydrological forecast the Zhinvali Water Reservoir released 600 m/s through
the spillways in order to regulate the extreme flow and to cut the peak of
the flood wave. There were damages along the Aragvi and Mtkvari Rivers
because of the high spill flow but the Zhinvali Reservoir safely operated and
degreased the peak flow.

The flood forecasting is very important for preparing the Zhinvali Reservoir
for regulating the high flows. Releasing the waters from the reservoir for
creating empty volumes needs time therefore long- and middle-term
forecasts are needed. The spring high flow is forecasted based on snow
masses in the mountains. However, the melting process is strongly affected
by the temperature regime and rainfall events occurring during the period
March-May. Therefore updating of the forecast should be done continuously
based on short-term weather forecasts as well as the meteorological
observations.

West Georgia

Air masses from south-west were transferred at the end of December, 1986,
which caused have precipitation. The snow cover depth reached 4-5 m in the
south slopes of Great Caucasus in the third decade of January 1987, i.e.
significantly exceeded the average values. Then rain showers hit the region
and contributed to the intensive snowmelt. As a consequence the water level
in the rivers rose rapidly (4-5 m in a few hours) and Bzifi, Kodori, Enguri,
Khobi, Tekhura, Abasha Rioni, Kvirila, Dzirula Rivers conveyed extremely
large water volumes. River Rioni exceeded its earlier historical maximum
water discharge because its peak flow was as high as 4850 m3/s. The size of
the inundated area on the Kolkheti Lowland reached 200 km2.

The losses were severe because 150 people died. Material damages reached
nearly USD 700 millions. The flood destroyed inter alia 3150 houses and
2150 objects, 16 km railway lines, 1300 km roads and 1100 km power
transfer lines. (attention: these figures are not in the

The forecasts of extreme high flows induced by snowmelt are based on the
same methodology as in the Mtkvari River Basin, however, the rainfall
generated runoff has more influence on the flow regime both in terms of
frequency, impact and geographical extent. That means that the potential
lead-time of the flood warning is short even on the downstream river
reaches.

Flash floods

The character of these events is different from the high spring flow because
the flash floods occur suddenly, in turn, the forecasts lead-time is short:
32 of 59

minutes and hours in the upstream regions and days in the downstream
parts. The forecasting method in the upstream parts are based on rainfall-
runoff functions and on the downstream reaches on statistical relations
between river gauges located in the upstream basin vs. on the lower river
reaches.

The main difference between the spring floods resulted by snowmelt and the
flash floods induced by rainstorms is the wide difference in the lead-time.
(Lead-time is the length of the period the forecaster can warn ahead about
the approaching extreme event.)

Flash floods can hit almost any part of Georgia, however, the most
vulnerable regions are Alazani River, left bank, Rioni River Basin, Inguri
River Basin. Heavy flash floods occurred on 11-12 1987 in Samtskhe-
Javakheti and Shida (inner) Karti regions. The return probability of
maximum flows was less than 1% in rivers Kura ((in Likani), Liakhvi (in
Kekhvi), Tetri (white), Aragvi (in Pasanauri), Paravani (Khertvisi). The losses
were as much a 1.4 million USDs.

* * *

In summary, we can note that the conditions creating floods in Georgia are
as follows.

(i) Intensive snowmelt together with large amounts of rainfall
(ii) Heavy rainstorms during summer / autumn seasons
(iii) Large territorial frontal rains during the autumn time
(iv) Intensive rainfalls during the winter season in the coastal zone of
Black Sea.

The probability of flood events is high almost in the entire territory of
Georgia, however, as the NEA Hydrometeorological Department assessed
that the most vulnerable areas were (Figure 13):

- Mtkvari River Basin: districts Gori, Kaspi, and Tbilisi
- Alezani River Basin: left bank (Great Caucasus)
- Rioni River Basin: downstream reach, downstream of Kutaisi
- Inguri River Basin: district Zugdicki.

If the rainfall intensity reaches or exceeds the threshold values then special
attentions should be paid to the location where the storm occurred because
there is a high risk for flood flow. In case, the field surveys showed that large
amount of movable debris accumulated in the river catchment/s where the
rainstorm occurred then mudflow hazard is also high. Therefore is necessary
the regular survey of debris accumulations.



33 of 59



Figure 13 Flash flood hazards in Georgia.

Mud flow

The National Environmental Agency has prepared geographical zoning of
mud flow maps in Georgia (Figure 14). There is no methods available for
predicting concrete mudflow events but there are criteria helping delineate
the hazardous land sections where stone and soil materials are
accumulated, which can be moved by water masses. There are critical
threshold values for rainfall depths. When the rainfall intensity exceeds
these values, the probability of mudflow occurrence becomes high.

The critical precipitation heights per 12 hours (threshold values) having
potential causing extreme floods an in turn mud flow in the upstream valleys
are as follows

(i) >130 mm in coastal zones of Western Georgia
(ii) >100 mm in central and western part of Colchis lowland and
adjacent mountain slopes
(iii) >80 mm in the remaining part of Western Georgia and southern
slopes of Great Caucasus
(iv) >60 mm in remaining Eastern Part of Georgia


34 of 59


Figure 14. Mud flow prone areas in Georgia.


Recommendations for Key Element 1 (Knowledge Base)

The Ministry of Environment Protection and Natural Resources of Georgia,
incl. the National Environmental Agency is in command of knowledge of
different natural risks, their occurrence regions as well as the priority
objectives of the protection. These zones were marked based on processing of
observations as well on extreme events having taken place in the past.

Today, the delineation of natural hazard zones is continuing by
implementing GIS technique because the available information on protected
objectives vs. character of natural risks provides a first option for preparing
preliminary emergency action plans. The maps are very useful instruments,
however, the scale (1:50.000) is not detailed enough for preparing plans for
risk assessments. Concerning the central subject (floods) of the current
paper, the Consultant Team suggests to develop the flood maps for the main
flood prone areas based on maps of larger scales and using hydraulic
computations for assessing the inundation areas.

The flood mapping work requires lot of data and working time therefore
efforts must be focused on priority areas. As the Hydrometeorological
Department assessed that the most vulnerable areas for floods were

- Mtkvari River Basin: districts Gori, Kaspi, and Tbilisi
- Alezani River Basin: left bank (Great Caucasus)
- Rioni River Basin: downstream reach, downstream of Kutaisi
- Inguri River Basin: district Zugdicki.

35 of 59

Therefore the flood risk maps should be prepared first for these areas. The
phases of these works include hydrological assessments to reveal the
hazards (size and frequency of floods), recording the human groups, objects
(houses, industrial objects, infrastructures) in the flood prone areas, GIS
processing (creating DTM, locating settlements and objects). When the GIS
layers of hydrographical and morphological data as well as the DTM are
ready, the hydrological data has been collected, the hydraulic parameters are
defined then the hydraulic computations can be made. The results will be
inundation areas at different flood flow levels. Detail of the work are
presented in Appendix 3.

The second recommendation concerns the assessment of mud flow hazards
and risks. The mounted material in the mud flow prone valleys should
reassessed in turn the hazards and risks should be updated according the
current conditions.

5.3 Monitoring and Warning Service (Key Element 2)
5.3.1 Organization of the Hydrometeorological Department
The National Environmental Agency, Ministry of Environment Protection and
Natural Resources of Georgia is responsible organization for issuing hydro-
meteorological forecasts, including flood predictions and warning. The core
of the forecasting service is the Hydrometeorological (hereinafter Hydromet)
Department with a staff approximately 180 persons grouped in different
divisions chart in Figure shows.

For clear understating of the functions of different units we briefly describe
their tasks:

- Hydromet Web Section takes care of the web site (3 persons)
- Climate Section focuses on the studies related to global
climate change (4 persons)
- Meteorological Section collects, processes and stores the
meteorological observations, which are transferred from the
Observatories on monthly base (10)
- Hydrological Section collects, processes and stores the
hydrological observations, which are transferred from the
territorial centers, so-called Observatories on monthly base
(4 persons)
- Division of Hydro-meteorological Hazards and Damage
Mitigation includes two sub-divisions such as recording the
potential natural hazard locations and planning mitigations
measures (8 persons)
- Telecommunication Service collects the operative (daily) from
Observatories, processing the information from the mezo-
scale meteorological forecasting systems and distributes via
emails the Daily Hydro-meteorological Bulletin. The Service
36 of 59

has two groups such as information service and R&D (13
persons)
- Hydro-meteorological Forecast Bureau (14 persons) which
includes three sub-divisions such as (1) Weather Long-term
Forecast Service, (2) Weather Short-term Forecast Service and
(3) Hydrological Forecast Service (12 persons)


Figure 15. Organization pattern of the National Environmental Agency.

5.3.2 Monitoring Service

In the paragraphs below highlight the situation of steady monitoring network
in the main regions, which have been targeted by international support for
37 of 59

the recent years. Then we will look at the status of hydrological field surveys
in todays Georgia.

Network

In general, we can note that significant reduction has happened in the
hydrological observation network in Georgia during the transition period.
The number of functioning hydrological stations has been 67 during the
period of 1930-1990, while today only 21 (according other sources 30)
gauges are reporting. It is far below the necessary level.

There are plans to increase and to improve the observation network by
adding 15 new (reconstructed old) gauges. in 2010, which plans can be only
considered as first step in the rehabilitation of the hydrometeorological
network.

There are about 45 observation points left from very well developed
meteorological network that existed until the late 1980s. From this number
currently only 13 stations can satisfy international standards defined by
WMO. From the next year the agency plans to purchase and install
additional 7 fully automated stations for measuring all major meteorological
parameters and 15 stations for recording selected weather parameters.

Current Status of the upgrading of observation network in main river basins

1.1. River Alazani:

River Alazani (Figure 16) does not directly threat settlements but the
tributaries do where the runoff concentration time is around 30 minutes.

- Floods and mud/stone flows on the left-side tributaries,
which take their sources from the North on the Caucasus
slopes.
- Mud and stone flow on right side tributaries, which river
beds are dry most of the time, take their sources on the
slopes of Tsiv-Gombori Mountains as a result of rainstorms.



38 of 59


Figure 16. Alazani River Basin.

The above tributaries are ungauged while automatic water level recorders
were funded by USAID and installed at Chiauri, on the River Alazani but not
in function because of some technical problems. One semi-automated
station, installed also by USAID, is functioning at Shakriani (near Telavi),
measuring water levels only.


1.2. River Rioni

Combination of intensive snowmelt and rainfall can cause sever floods along
the Rioni River. The water level in the tributaries of Rioni River can increase
several meters in a few hours. Such event happened in January 2006 when
the old historical maximum water levels and discharges were exceeded.. The
downstream part of the basin is a flat flood plain (prone) area. The Figure
Figure shows computed inundation territories modeling propagation of flood
waves of 1%, 5%, 10% and 20%

The water courses in the Rioni River Basin have sparse observation network
and the upstream basins are ungauged today. The Finish Government and
CIDA funded procurement and installation of automated stations for water
level and precipitation measuring. The stations have been installed on river
Rioni and tributaries (Tskhenistskali) but are not in function yet for
technical reasons.

39 of 59


1.3. River Mtkvari

The Kura River originates in the Kizil-Giadik mountain range in Ardahan
province in Northeast Turkey at the altitude of 2740 m, winding its way
through mountainous regions in Turkey. Tributaries flow northwards from
Armenia and join the Mtkvari in Georgia. From the source area to the city of
Borzhomi the river bears mountain character, after this point lowland regime
prevails. The flood flow regime is reflected by geographical location of the
catchment areas. Flash floods often hit the upstream basin following the
rainstorms while the lowland reach is mainly flood prone in case of long-
term water feedings from intensive snow melts accelerated and amplified by
areal rainfalls.

Automatic water level recorder is installed on the River Kura at Red Bridge
and at Sadakhlo on the River Debet, which a main tributary (on the left side)
of the River Kura. The recorders are not in function today because of some
technical problems. The other rivers are practically ungauged today.

5.3.3 Data collection

Primary data

The daily operational observations and monthly journals are collected from
the field stations by the territorial centers, so-called Observatories and
forwarded to the Hydromet Department. The data from the stations
transferred in WMO-codes via cell phone, fax and internet in case of
available.

The Telecommunication Service obtains the data from regional weather
forecasting systems and also collets the daily operational observations from
the observatories and transfers to other units at Hydromet Department.

Processed data

The observation data after checking is forwarded by the Observatories to the
Hydrological Department. The telecommunication means depend on the
technical availability.

Internal data flow

There is no database manageable and accessible via local IT network at the
Hydromet Department. The data transfer is still manually made although it
is made in digital form or on paper. The integrated database within the NEA
as the first presumption for the Integrated Early Warning of Natural
Disasters in Georgia.


40 of 59


5.3.4 Field works

Only limited field works are performed in Georgia today while there is a
significant need for measuring water discharges, to perform hydro-
morphological investigations and to carry out catchment prospecting. The
problems stem from inadequate equipment base and lack of manpower.
Underlying factors are insufficient financial resources, lack of donor
coordination (a number of transferred equipment and software tools are not
in use), missing competence in installing, operating, utilizing and
maintaining of the donated equipment.

The snow surveys should be performed between 25
th
February and 20
th

March because the data is extremely important for issuing flood warning of
the spring high flows. This field activity has also significantly been reduced
during the transition period. The snow survey started in 1955 and included
35 routes located in the high mountains. Today only maximum 20 snow
routes are surveyed, which programme does not provide enough information
for preparing the flood warnings for the high spring flow. Remote sensing is
not yet used for snow resource estimations.

Recommendations for the monitoring service (Key Element 2)s

The availability of the hydro meteorological observation data is the basic
assumption for any early warning/prediction activity. The primary
observation data needed for flood warning / prediction are as follows
precipitation, snow resources, soil conditions, temperature, water level and
discharges in the rivers; secondary data such as evapotranspiration,
moisture content in air (relative saturation), wind.

The observation data is produced in different ways such as by traditional
visual observations, automatic gauges stations, radar surveys and remote
sensing. The accuracy, territorial density and timely frequency of the
observations are a determining factors of the early warning systems, i.e. the
more observation stations, the more frequently the measurements are made,
the better accuracy of the parameters achieved we can provide the more
reliable forecasts of the extreme flood events.

Along the above observations there is a need for hydrographic and
hydromorphological parameters, soil characteristics and relief data (maps),
which data produced by field investigations like hydrometric measurements,
geodetic mapping, soil studies.

The installation and operation of the hydro-meteorological networks are
expensive and the field works are also resource demanding, therefore a kind
of optimizing procedure is necessary when designing an observation systems
and planning the field campaigns. The investment, running and operational
41 of 59

costs are on one side vs. priority factors requesting the flood warning /
prediction are on the other side. The latter one are as follows

- human life should be protected (of highest priority)
- important social services should function (health care)
- transport function, supply of population (food, water energy
etc.) should be maintained
- damages of properties should be minimized

The design of the observation network approaches the optimal system from
two stand points such as

- the objectives (settlements, industrial plants, energy
sources etc.) to be protected vs.
- the character of the threats, which are spring high
flows, flush flood, mud and stone flows in our particular
case.

We have to keep also in mind that the basic and timely information for the
early warning systems stem from both steady network and field campaigns
and both components are equally important.

Monitoring of ungauged rivers

This river catchments are first, second, third etc. tributaries to main
streams. The runoff concentration time is very short (from 0.30 min to a
few hours) in these catchments. Therefore visual observation of river gauges
made on daily base would not provide any useful information for flood
warning at these locations. On the other hand, installation of automatic
gauges for measuring and transmitting the water levels is hardly viable in all
these catchments having high potential for flash floods. The regional weather
forecast models cannot provide either data in such fine timely and spatially
resolution, which would give reliable information for issuing flood warnings.
The only solution is to assess the potential hazards in the ungauged
catchments and propose installation of automatic raingauges where it is
required and feasible. The location is an issue of multi-variable
consideration, namely on one hand the source areas should be monitored on
the other hand the station should be accessible for maintenance and for
energy supply safely even in extreme weather conditions (that is the main
point).

During the Soviet time there was weather radar network in function in
Georgia, which was used for hail prevention, storm tracing, rainstorm
forecasting. The Division of Hydro-meteorological Hazards and Damage
Mitigation has prepared a new map for installation of weather radar network,
which would cover the major part of the country. The Consultant proposes
that this option should be analyzed both in terms of effectiveness and
feasibility.

42 of 59

Monitoring gauged rivers

The visual observation of river and rain gauges is still a practical option in
the country. The perspectives of rehabilitating the abandoned gauges must
be considered. The observation should be made two times per day (as it was
done before) as long as the flow stays under alert level. The observation
should be made four or more times as soon as alert level exceeded and every
two hours above preparedness level.

Development of telemetry should be continued, however, first the already
purchased automatic stations should be operation.

Field Campaigns

The snow surveys must be raised to the level they were practiced before the
transition time. It means that frequency of the measurements should be
increased and the routes must cover necessary locations in terms of length
and altitude. The equipment and manpower of snow survey teams must be
upgraded. The remote sensing should be taken as additional data source for
estimating the snow resources. Therefore, such R&D study must be initiated
and competence development shall be planned in this field.

The basic hydrological parameter is the water discharge (m
3
s
-1
). These
measurements require skilled personnel and are also time consuming. In
addition, the works should be timed according the hydrological events, i.e.
do not fit into the regular working hours. Therefore discharge measurements
are very rare in Georgia today.

The old current meters are outworn and not-calibrated. Two Doppler type
flow meters were obtained via donor programme. However, one of them is not
functional today. The reason for failure is not clear for the Consultant yet. In
any case, both instruments must be in function and then it is possible to
look into further upgrading of the equipments of the field teams.

Data management

The Consultant recommends an IT-based integrated information system in
order to make effective data flow between different units. The goal of the
development work is to speed up the data flow and processing, to make safe
the data exchange and storage between units of the Hydrometeorological
Department as a first step.

The following paragraphs will describe the status of the hydrological
prediction activities at the NEA.

5.3.5 Prediction Service

43 of 59

The hydrological prediction activity divided into two separate groups because
generic differences in sources of the water masses, in observation methods,
in forecasting tools and in lead-time. The two groups are such as

(i) spring floods are induced by snowmelt, which
can be amplified be rain and
(ii) flash floods resulted by rainstorms.

Forecasting of the spring floods

Regression equations were established based on hydrological and
meteorological time series observed between 1970-1980. The parameters in
the regression equations were set permanent earlier but today are regularly
updated using computer technique.

Input parameters are as follows
- water equivalent in snow cover
- temperature regime forecasted for the snowmelt period
- water content in the upper soil layer measured before the snow cover

The spring flow forecast with longest lead-time covers three-month period,
i.e. the first forecast given at the end of March and stretches to the end of
June. The prognosis is a moving window and updated each month, i.e. in
April and May. The updating of the snowmelt prognosis is made using the
long-term forecast of air temperature taken from mezzo-scale weather
forecast systems (described below). In case of heavy rainfalls during the
snowmelt, extraordinary forecasts should be prepared and issued because
these situation carry particularly high risk for disastrous floods.

The National Environmental Agency is working on an up-to-date approach,
i.e. on adaptation of the outputs (predicted temperature regime and
precipitation values) from regional weather forecasting systems. The time
resolution of these schemes is 6 and 12 hours, while the spatial resolution
scale is of 14x14 km.

The weather forecast based on two regional systems such as WRM-EMS (US)
and HRM (GE). The size of the resolution cells is 14x14 km. Forecasts of
precipitation and air temperature are necessary in order to increase the lead-
time of flood warnings and forecasts in a river basins in Georgia.

The Weather Research and Forecast Environmental Modeling System (WRF
EMS) is a complete, full-physics, state-of-the-science numerical weather
prediction (NWP) system that incorporates dynamical cores from both the
National Center for Atmospheric Research (NCAR) Advanced Research WRF
(ARW) and the National Center for Environmental Predictions' (NCEP) non-
hydrostatic mesoscale model (NMM) releases into a single end-to-end
forecasting package. All the capability of the NCEP and NCAR packages are
retained within the WRF EMS; however, installation, configuration, and
running of the NCEP and NCAR versions have been greatly simplified to
44 of 59

encourage its use by operational forecast offices, private industry, and the
University community.

The High resolution Regional Model (HRM) is a flexible package for
Numerical Weather Prediction (NWP). The Deutscher Wetterdienst (DWD)
provides this comprehensive package to meteorological services, universities,
and research institutes world-wide. This model uses the topographical data
sets for any region of the world at mesh sizes between 30 and 5 km which
are prepared at the DWD on request. The HRM model is used as NWP model
at meteorological centres in many countries. Generally, the output of this
model is in good agreement with the observations.

Short-term prognosis floods

The number of forecasted stations radically reduced from 161 targeted
earlier to 21 predicted today. The forecasting group issuing the short-term
forecast has lost major part of its manpower. There were 14 specialists in the
office while only three persons are working there today.

As it mentioned above the operative forecasts are made for too few stations
and exceptionally based on intuitive assumptions. Some information sources
say that either mathematical formulas or graphical tools were never used for
short-term daily flow (water level) predictions. There were graphical tools
developed by excellent specialists for average flow forecasting (10 day-
average etc.), which values were necessary for water reservoir control and
planning of hydropower production.

* * *

The NEA has obtained via donor support the Delft-FEWS software package
and one person at the Hydrometeorological Department has been shortly
trained on setting up HEC HMS and configuring the Delft-FEWS platform.

A small pilot application of Delft-FEWS was build in the Rioni basin, which
work was supervised by international expert from Deltares (NL). The flood
early warning system used HEC HMS as hydrological model, and the inputs
were taken from limited area weather model WRF (Ucar) and HRM (German
Weather service) as input. The hydrological model has been build on the
basis of a digital terrain map with 90x90 m resolution published by the
CGIAR consortium. The topographic analysis allowed to extract subbasins
and river channels (channel lengths and slopes). The routing scheme is
based on Muskinhum-Cunge schematization, with assumed river geometries
and roughness.

The model has been calibrated on digitized precipitation and temperature
data from the period between 1980 1992.


45 of 59

The problems encountered during the Delft-FEWS adaptation are, inter alia,
lack of river cross sections, outdated discharge curves, missing time-series
after the Soviet time, and not enough training was given.

The second important issue, which is waiting for solution is the online data
input from the automatic hydromet gauges into the Delft-FEWS. As the
technical specification of the Delft-FEWS states that this option is provided,
however, the technical details is not clear at the moment and no action was
taken into this direction. In addition, the automatic stations do not forward
online data into the Agency because of some technical problems, which
might be soon solved.


Recommendations for the prediction activities (Key Element 2)

The next step of development at the Hydromet department is under
consideration, namely the downscaling of outputs from the Regional Weather
Forecast Systems because the current timely and spatial resolutions do not
allow the predictions / warning in catchments of smaller size, which do have
high potential flood hazards and therefore there is a need for extreme flow
prediction at these locations. The downscaled cells would have spatial
resolution of 7x7 km.

The Consultant proposes competence development of the personnel involved
in this R&D work. The most effective way for the competence development to
ensure a couple of months study visit at the mother institutes running the
models: NCAR or NCEP in USA or DWD in Germany or other places where
the coupling of these weather forecasting systems to hydrological prediction
models successfully were successfully implemented.

The Consultant sees necessary the upgrading the short-term forecasting
activities. There are relatively simple equations for real-time prediction with
adaptive parameter updating, which are able to predict even such rapidly
changing water regime in Georgia. In turn, the quality (accuracy) and
frequency of the forecasts could be improved. In addition, more station could
be forecasted even by limited number of personnel.

Of course, we have to strive after limited number of models to be
operationally used but there is no sense to look after general model package,
which is fits to all rivers and to all cases in Georgia. Instead, there is a need
for method selection, where on one hand the priority water objects should be
chosen on the other hand the data availability should be realistically
assessed.

The Consultant sees that there is a need for strengthening the personnel
engaged in the operational forecasting / warning activity: manpower must be
extended. A mixture of experienced specialists and IT-skilled personnel
would be favorable for the upgrading of forecast services. Rather more
because the coupling procedure for transferring data from the numerical
46 of 59

forecast of meso-scale meteorological models HRM/WRM-EMS into forecast
hydrological model Delft-FEWS has been initiated. The testing procedure has
been in process for three months. The work should lead to a coupling of the
meteorological parameters from HRM/WRM-EMS to the operational flood
forecasting and warning tools (model/s). However, the coupling process is
moving slowly. The reasons may be several inter alia lack of supervision,
personnel occupied by other tasks, low interest from operational side for
using this technique in the practice. The Consultant proposes to intensify
the adaptation process, first to check the obstacles hindering the operational
implementation and take measures to fulfill the objectives set for this work.

Another, major point is setting-up online (or semi online) prediction/warning
system. There is not any sene to install expensive automatic gauges if they
are not on-line connected to models capable to issue forecast in real-time
mode. The Consultant suggest to make parallel efforts, i.e. to install the
automatic gauges within a short time but during the installation period to
set-up a hydrological model capable to issues online flood predictions

The next step for achieving forecasting stage of the HMS model is that the
module needs to be interfaced with real-time observations from the
telemetric stations.

5.4 Dissemination and Communication (Key Element 3)

National Environment Agency completes Bulletin and publishes Bulletin on
daily base. Bulletin a table with 21 river gauges:

Alert water level Observed level
today
Forecasted water
level lead-time 24
hours
Forecasted water
level lead-time 48
hours




The Bulletin is sent in hardcopies to President Administration, Security
Council, Ministries, operators of Hydropower Plants, and other users upon
requests. In addition, the Bulletin is distributed via e-mail. Information is
also given to municipalities in case of extreme events.

NEA units forward operational warnings for natural disasters to Directorates
of Road Transport and other authorities. This information is directly
forwarded from the NEA units to the end-users by phone and fax. These
services are often based on special contracts made between the NEA and the
end-user.

Recommendations for dissemination and communication (Key Element
3)

47 of 59

One of the main objectives of the Study is promoting the establishment of
Integrated Early Warning System of Natural Disasters. The basis of the
natural disasters is so much different therefore the methodological
approaches to foresee them also special for them. In addition, the forecast
term has different meaning for the flow regime in rivers, rainstorm,
windstorm vs. avalanche, mudflow and landslide. For example, the flow
regime can be forecasted with acceptable accuracy, which means that we
can say that at certain location what the water level will be and when. While
in case of the other natural catastrophe events we only can say that the
probability of occurrence is high or very high in the current conditions.

Therefore the intersection of the Early Warning System of Natural Disasters
is the integrated database and the integrated dissemination system. In turn,
the Consultant Team sees necessary to solve these two tasks within the Key-
element Communication and Dissemination.

Developing effective communication strategies has two aspects; the technical
aspect relates to build or strengthen robust hazard-resistant communication
systems; the institutional aspect relates to the maintenance of relationships,
i.e. the need to establish and maintain effective links and working
relationships among the actors involved in the early warning communication
chain. The current section deals with the hardware component, i.e. makes
recommendations for building the integrated information system while the
next section focuses on institutional links between the actors involved in the
natural disaster mitigation. Two information systems should be developed,
namely, the Integrated Database at NEA and Integrated Early Warning
System in Georgia. The work for these interrelated system should start with
the system design, which will create underlying documents for the decision-
makers concerning needs for programme and hardware tools.

Integrated database at NEA

An integrated database should be build at the NEA, which work can be done
in several phases. First, the Hydromet Department should integrate the data
arrays already saved on magnetic holders. The integration should start with
a data inventory and a system design. Based on the system design and data
volumes can be decided the hardware needs of the system.

The integrated database will be open only for the Ministry of Environment
Protection & Natural Resources (MoE) of Georgia and contain of two modules

Historical data a.o.

(1) Runoff data from stations
(2) Gridded meteorological data calculated from synoptic
stations
(3) Hydrological, meteorological and related data from the
special studies, for example flood inundations; flood
caused mudflow etc.
48 of 59

(4) Hydrological data such as snow equivalent of water,
soil moisture, information on ground water, ice
observations, lake levels, lake and river temperature
(5) Information of earlier avalanches: location, impacted
area, velocity, head power, volume, size, initial
conditions (snow depth)
(6) Mud flow prone areas: locations, mounted material,
which was moved, impacted areas, rainstorm
intensities in the region when the mudflow occurred
(7) Landslides: locations, geological parameters, impacted
areas.

The historical database will be continuously developed and detailed
information on the current contents will become available.

Operational observations a.o.

(1) Runoff data from stations
(2) Gridded meteorological data calculated from synoptic
stations
(3) Hydrological, meteorological and related data from the
special observations
(4) Hydrological data such as snow equivalent of water,
soil moisture, information on ground water, ice
observations, lake levels, lake and river temperature
(5) Information of risk for avalanches: location, potential
impact areas, snow depth when the risk for avalanche
can be very high
(6) Mud flow prone areas: locations, mounted material,
potential impact areas, rainstorm when the mudflow
risk can be very high
(7) Landslides hazards: locations, geological parameters,
potential impact areas.

It is recommended to build the database on GIS platform and use GIS
technology advances achieved over the past several years. They provide
effective and efficient data management for processed hydrographic data.
The geodatabase model fits well to database designs and applications
associated with the collection, documentation, distribution, and analysis of
large amounts of vector, raster, and surface modelling data. It also provides
the end- user with the options of setting states and behaviours for each data
object. In our case, it means that not only relief and morphological data can
be stored in the database but also hydrometric measurements and hydro-
meteorological observations (time-series), geophysical parameters, soil
characteristics, snow conditions etc., which the natural disaster warning
system needs. The later ones will be assigned to the geographical data
objects. Using this approach provides the means of storing, analyzing and
modelling data and enabling the existence of intelligent data objects and
methods to distribute value added hard and soft copy products to end-users.
49 of 59


Integrated Early Warning System

The Consultant Team recommends a server based Early Warning System.
When an client sends a request to a server, the server processes and
responds to the request. Typical requests generate maps or retrieve
geographic data, objects at risk or protection measures etc. for a given local
extent. The hierarchy of the information is outlined below.

(1) Information from Hydromet to Central Organizations

In case of emergency, the hydromet information must be issued frequently.
The content is observed and expected water levels, comparison with alert
levels, confidential interval, flood duration. Consequence analysis: impacted
rivers, regions, districts, settlements, inundation maps and expected
measures taken operators of the hydropower plants.

(2) Information from Central Organizations to Regional Emergency
Centers

In case of emergency, the preparedness information must be in separate
break-downs for the Regional Centers command area and issued frequently.
The content is observed and expected water levels, comparison with alert
levels, confidential interval, flood duration. Consequence analysis: impacted
river reaches, districts, settlements, inundation maps, list of objects and
infrastructures at risk, secondary impacts (wash-away of pesticides etc.)
expected measures of the hydropower plants (sudden releases of waters etc.).

Requested emergency measures inter alia: protection activities, relief
operation, resources necessary (manpower, machinery, construction
materials, etc.), list of organizations and operators to kept on permanent
contact.

(3) Information from Regional Emergency Centers to the Municipalities

In case of emergency, the preparedness information must be in separate
break-downs for the Municipalities command area and issued frequently,
the content is water levels observed and expected, comparison with alert
levels, confidential interval, flood duration; consequence analysis: impacted
river reaches, districts, settlements, inundation maps, objects and
infrastructures at risk, secondary impacts (wash-away of pesticides, waste
waters, landfills etc.) expected measures of the hydropower plants (sudden
releases of waters etc.).

Requested emergency measures inter alia: protection activities, relief
operation, resources necessary (manpower, machinery, construction
materials, etc.), list of organizations and operators to kept on permanent
contact.

50 of 59

(4) Information from Municipalities to the Responsible Persons
(inter-municipality division)

In case of emergency, the preparedness information must be issued in
break-down for the responsible persons and the communication should be
permanently kept; the content is houses, objects and infrastructures at risk,
secondary impacts (wash-away of pesticides, waste waters, landfills etc.)
expected measures of the hydropower plants (sudden releases of waters etc.).

Requested emergency measures inter alia: protection activities, relief
operation, resources necessary (manpower, machinery, construction
materials, etc.), operators to kept on permanent contact.

The Municipalities as well as the responsible persons should inform the
public and mobilize them in case of necessity.

(5) Back-to-Office Reporting

The information chain should also function in a reverse direction, i.e. reports
must be made on the current situation in the field, on taken measures,
requested support, etc.

A main issue is to decide the operator of integrated early warning system.

5.5 Response Capability (Key Element 4)

Response Capability is perhaps the most important and vital link among the
four elements of the people centered Early Warning System discussed above.
As it is well defined the ultimate goal for any early warning system
established even on the latest possible technology and widest possible
monitoring network is first of all to protect human lives, households, and
public goods from devastating results of natural or manmade disasters.
Accordingly, if this last element, responsible for immediate response
activities and preparedness of communities to address any such challenges
is missing, or does not work properly, the entire system cannot fulfill the
goal.

This is true for Georgian situation as well, where the fourth element mostly
is missing. Currently Emergency Management Department (EMD) is the only
institution that more or less satisfies disaster preparedness and response
planning requirements of the Key Element-4 on a national scale. However
this is only one part of the entire Response Capability. Other issues, like
Community Response Capacity, Public Awareness and Education, Public
Perception for Natural Hazards on local and at the same time most
vulnerable levels are missing. To improve the situation on a short run EMD
should take more active role before the planned Regional Emergency
Management Service Centers (REMSCs) will be established under the
Ministry of Regional Development and Infrastructure. It is believed that these
51 of 59

Service Centers will fill the gap between the national and regional/local level
response activities, as far as intensive training, procurement and technical
assistance is foreseen to strengthen their capacity. Taken as this agenda of
the Georgian Government will be implemented sooner, the Response
Mechanism in Georgia may look the following way:




Figure 17. Immediate Response Mechanism of the FEWS in Georgia.


It is also equally important that the institutions responsible for creating
knowledge base, and modeling and interpreting scientific data, will develop
hazard and vulnerability maps for different types of disasters (floods, flash-
floods, as well as other climate related disasters) customized to the regional
and local conditions. These data later will play central role for creating
targeted disaster response plans for every vulnerable locations on a
municipal and community level.

Schematic flow of the proposed initially processed and modeled regular data
is shown bellow (Figure 18). This scheme relates only to a filtered and
interpreted data that assists appropriate agencies define scale of
preparedness in case of increased water levels. The list of the agencies
receiving the data will include: the National Security Council; Emergency
Management Department (EMD) of the Ministry of Internal Affairs; Regional
52 of 59

Emergency Management Service Centers (REMSCs) under the Governors
Offices; Ministry of Environment Protection and Natural Resources (MEPNR);
Ministry of Health (MOH); Ministry of Economic Development (MED);
Ministry of Agriculture (MOA); Ministry of Foreign Affairs (MFA); Ministry of
Defense (MOD).




Figure 18. Primary distribution network of the FEWS in Georgia.


One other important step for establishing proper Response Capability is to
increase public awareness and confidence building for warning services on a
community/municipal level, by actively engaging: local media, TV/radio
broadcasters, NGOs, schools, volunteer organizations, religious and political
leaders in the annual planning, evaluation and training processes. It is
absolutely necessary to update contingency and evacuation plans annually
to address new circumstances related to the rapid climate change and thus
altered hydro-meteorological information. And finally, communities and local
authorities should be trained to recognize different levels of the warning
signals, as natural hazards are not always fully predictable and falls alarms
may build wrong credibility for the system in general. One example of the
three different levels of preparedness in case of flood warning for local
community at the Binahaan river basin, Philippines (study by Olaf Neussner,
53 of 59

Allen Molen, and Thomas Fischer, GTZ, 2008) is shown on the figure bellow
(Figure 19). In this example local community has detailed instruction for
each level of warning staring from simple alert to the preparation and then
evacuation. Well established contingency plans in edition should have
detailed evacuation scenarios and exact transportation routes, including
roles to be played by each municipal / regional sectoral unit and rescue
services. And finally, the plan should include detailed description for post-
disaster relief operations coordinated again by local/regional authorities and
disaster management teams.




Figure 19. Flood Warning Plan, Binahaan river basin, Philippines.


6 Conclusions

Based on the request of the Ministry of Environment Protection & Natural
Resources (MoE) of Georgia, UNDP assists the MoE in conducting a needs
assessment and a feasibility study for establishing an early warning system
for natural hazards in Georgia. The study was implemented in close
collaboration with the UNDP Disaster Risk Reduction Advisor, the UNDP
Environment & Energy portfolio, MoE, National Environmental Agency and
other concerned national authorities.

The Study focused on the flood warning subject but the prospect of the
integrated early warning system was also kept in view throughout the
project. The National Environmental Agency, Ministry of Environment
Protection and Natural Resources of Georgia is responsible organization for
issuing warnings and forecasts for natural disasters (except the earthquake).
Therefore, the Consultants concentrated on this Agency concerning issuing
the warnings but other R&D organizations were also screened. The response
54 of 59

to the warnings is controlled by other organizations, which were also met
and their woks analyzed.


The Conclusions briefly summarize the findings and the recommendations,
while the needs of specific tasks in terms of equipment, staff and
competence development are included in Appendix 3.

The Project tasks as well as the current Report followed the structure of the
Checklist of the Early Warning Systems published in Proceedings of the Third
International Conference on Early Warning (Bonn, 2006) and includes four
Key Elements

1) Risk Knowledge
2) Monitoring and Warning Service
3) Dissemination and Communication
4) Response Capability.

The Conclusions in a concise form will provide the findings and
recommendations in the same sequence.

Risk Knowledge

The studies of natural disasters have been done for decencies and performed
at high scientific level in Georgia a good knowledge base available to tackle
the natural disaster reduction in Georgia. The institutions show definitive
signs of recovery after the severe difficulties experienced during the
transition period. However, the shortage of funds is still obvious, which
result lack of personnel, limited access to modern equipments and reduced
monitoring programmes.

The NEA utilize the advanced technologies, for example computers, digital
data management, GIS, telemetry, modeling etc. However, the number
specialists skilled for advanced technique is limited therefore competence
development is needed and hardware and software base should be
completed. The assessment of inundation areas requests application of
advanced hydraulic and GIS technique therefore competence development
should be done at international referred institute and international
supervision is needed for the programme.

The first priority recommendation is that the flood mapping work must focus
on priority areas such as

- Mtkvari River Basin: districts Gori, Kaspi, and Tbilisi
- Alazani River Basin: left bank (Great Caucasus)
- Rioni River Basin: downstream reach, downstream of Kutaisi
- Inguri River Basin: district Zugdidi.

55 of 59

The second priority recommendation within the Key Element of Knowledge
Base concerns the assessment of mud flow and landslide hazards and risks
in Georgia. The mounted material in the mud flow prone valleys should
reassessed and in turn the hazards and risks should be updated according
to the current conditions. These works requires upgrading of the equipment
base and additional staff resources for field surveys.

Monitoring and Warning Service

Monitoring

In general, we can note that significant reduction has happened in the
observation network in Georgia during the transition period and the network
density as well as the measuring programmes are far below the necessary
level.

The availability of the hydro meteorological observation data is the basic
assumption for any early warning/prediction activity therefore the
rehabilitation of the observation network should continue and the field
surveys must be extended. Installation of telemetric raingauges is
recommended in the ungauged catchments where set-up of hydrometric
stations is not feasible but there is a strong need for warning of potential
flood and mud flow hazards. The Development of telemetry should be
continued in the main River Basins such as Mtkvari, Alazani, Rioni and
Inguri. But the Consultants, strongly recommend that first the already
purchased automatic hydrological stations should be put into operation.

The snow surveys must be raised to the level they were practiced before the
transition time. The equipment and manpower of snow survey teams must
be upgraded. The remote sensing should be taken as additional data source
for estimating the snow resources. Therefore, such R&D study must be
initiated and competence development shall be planned in this field.

The basic hydrological parameter is the water discharge (m
3
s
-1
) but the
measurements are very rare today. The old current meters are outworn and
not-calibrated. Two Doppler type flow meters were obtained via donor
programme. However, one of them is not functional today. The Consultants
recommend that both instruments were in function and further upgrading of
equipment base including instruments for hydro-morphological surveys.

Prediction Service

The National Environmental Agency is working on an up-to-date approach
when issuing the spring floods forecast while outputs (predicted temperature
regime and precipitation values) from regional weather forecasting systems
are used as input data in the forecasting model. However, the resolution
from the weather forecasts should be downscaled and therefore R&D work
should be done. The successful downscaling needs competence development
56 of 59

of a Georgian specialist at one of the internationally referred institute. In
addition, the hardware and software base must be significantly improved.

The number of forecasted stations has reduced radically because the
forecasting group has lost major part of its manpower and the observations
are not enough to predict water flow for the rest and important water gauges.

The NEA has obtained via donor support the Delft-FEWS software package
for short-term forecasting and one person at the Hydrometeorological
Department has been shortly trained on setting up HEC HMS and
configuring the Delft-FEWS platform. A small pilot application of Delft-FEWS
was build in the Rioni basin. The automatic stations purchased via donor
activities should be also connected to the Delft-FEWS but they do not
forward online data to the Agency because of some technical problems,
which might be soon solved.

The Consultant sees urgent upgrading needs of the short-term forecasting
activities. The upgrading requires actions in two directions. The personnel
engaged in the operational forecasting / warning activity must be extended
and a combinations of experienced specialists and IT-skilled experts is
desired. The second proposal is to intensify the adaptation process of Delft-
FEWS and HEC HMS. These works require competence development of the
personnel at international institute having long experience in the adaptation
works. In addition, international supervision is needed for the Delft-FEWS
and HEC-HMS adaptation at NEA. The hardware base should be upgraded
for operational use of these advanced techniques.

Dissemination and Communication

The main objectives of the current Study is promoting the establishment of
Integrated Early Warning System of Natural Disasters. The basis of the
natural disasters is so much different therefore the methodological
approaches to foresee them also special for them. In addition, the forecast
term has different meaning for the flow regime in rivers, rainstorm,
windstorm vs. avalanche, mudflow and landslide. Therefore the intersection
of the Early Warning System of Natural Disasters is the integrated database
and the integrated dissemination system. In turn, the Consultant Team sees
necessary to create the integrated early warning system within the Key-
element Communication and Dissemination.

Developing an effective communication strategy has two aspects; the
technical aspect relates to build or strengthen robust hazard-resistant
communication systems; the institutional aspect relates to the maintenance
of relationships, i.e. the need to establish and maintain effective links and
working relationships among the actors involved in the early warning
communication chain.

Two information systems should be developed, namely, an Integrated
Database at NEA and Integrated Early Warning System in Georgia. The
57 of 59

Consultants recommend preparing a system design, which will create
underlying documents for the decision-makers concerning needs for
programme and hardware tools.

It is recommended to build the database on GIS platform and use GIS
technology advances achieved over the past several years. The geodatabase
model fits well to database designs and applications associated with the
collection, documentation, distribution, and analysis of large amounts of
vector, raster, and surface modeling data.

The Consultant Team recommends a server based Early Warning System.
When a client sends a request to a server, the server processes and responds
to the request. Typical requests generate maps or retrieve geographic data,
objects at risk or protection measures etc. for a given local extent.

The Consultants provide an estimate of the integrated data management and
communication system at the NEA for issuing warnings of natural disasters,
however, the communication system of EWS for Georgia needs special
system design, which is far beyond the scope of the current study.


Response Capability

Response Capacity, the most valuable and vital element of the people
centered Early Warning System, is unfortunately the weakest and most
fragile part of a Georgian disaster risk reduction practice. Immediate disaster
response currently is limited by rescue operations by Emergency
Management Department (EMD) in case of the National Level disasters only.
Consequently EMD is a major player for a national Disaster Preparedness
and Response Planning, which however is absent on a regional and local
levels. Other important issues of the Response Capacity, such as Community
Response Capacity, Public Awareness, and skills to formulate proper Public
Perception on the Natural Hazards are fully absent on all levels, or lack
appropriate competence.

The Consultant Team recommends taking immediate actions to improve
and/or establish missing parts of the Response Capacity, especially on a
local/regional levels. Proposed by Georgian Government and planned to be
established early next year the Regional Emergency Management Service
Centers (REMSCs) will require intensive training and capacity building
programs, which on the other hand leads to immediate needs for adequate
financial and human resources that are not always easily available.
Fundraising and capacity building itself requires sufficient skills, time and
resources; therefore to fill the time gap the Consultants recommend
meanwhile utilizing EMDs resources more actively.

The Consultants recommend establishing formal network of the government
institutions receiving of filtered and interpreted early warning information on
a daily base that is crucial for scheduling immediate response activities and
58 of 59

contingency planning on institutional/sectoral and local levels. The list of
the agencies should include: the National Security Council; Emergency
Management Department (EMD) of the Ministry of Internal Affairs; Regional
Emergency Management Service Centers (REMSCs) under the Governors
Offices; Ministry of Environment Protection and Natural Resources (MEPNR);
Ministry of Health (MOH); Ministry of Economic Development (MED);
Ministry of Agriculture (MOA); Ministry of Foreign Affairs (MFA); Ministry of
Defense (MOD).

The next step for building proper Response Capacity in the country should
include actively engaging community leaders and local authorities at every
vulnerable location for developing local disaster preparedness and risk
reduction plans; involve local media, TV, schools, religious and social
organizations; train them, monitor, evaluate and update planning process on
a regular base.

59 of 59


7 Appendix

Appendix 1 Natural disasters and consequences in Georgia

Appendix 2 River water stages exceeding alert and flood levels in Georgia

Appendix 3 Cost estimates of upgrading the early warning system and
Competence development needs



APPENDIX 1
Intensity
(Amount)
Damage/
Hum.
Loss (mln.
GEL/Man)
Intensity
(Amount)
Damage/
Hum. Loss
(mln.
GEL/Man)
Intensity
(Amount)
Damage/
Hum. Loss
(mln.
GEL/Man)
Intensity
(Amount)
Damage/
Hum. Loss
(mln.
GEL/Man)
Intensity
(Amount)
Damage/
Hum.
Loss (mln.
GEL/Man)
Intensity
(Amount)
Damage/
Hum. Loss
(mln.
GEL/Man)
Intensity
(Amount)
Damage/
Hum.
Loss (mln.
GEL/Man)
1995 4 3,2/1 _ _ 2 0,5 666 132/4 693 96/1 8 3,2/2 7 12,7 247,6/8
1996 11 28,5/1 1,5 17,0 4 4,0/5 404 80,3 198 27,0/5 6 3,8/3 11 17,0 177,6/14
1997 12 38,0 2 26,0 3 1,0 510 102,0/3 318 44,0 10 4,2 14 35,0 250,2/3
1998 2 2,0/1 1 6,0 3 72,0/5 333 67,0/2 147 20,0/6 9 3,9/2 12 8,5 179,4/16
1999 8 30,5/1 _ _ 2 3,5 56 12,0/1 27 4,5 12 3,7/1 9 6,9 61,1/3
2000 2 2,0 6 300,0 2 1,0 65 13,0/1 23 3,0 7 2,1/1 7 5,8 326,9/2
2001 4 4,1 2,5 21,0 1 0,1 75 15,0 26 4,0 6 3,5/1 8 10,4 58,1/1
2002 16 78,7 _ _ 2 0,6 69 13,8/1 23 2,5/2 8 1,5 8 6,8 103,9/3
2003 6 4,2/2 _ _ 1 0,1 71 14,5/3 28 4,0 8 2,1/2 7 6,0 30,9/7
2004 10 20,5/1 _ _ 4 0,8 736 147,0/1 192 28,0 10 4,8/1 11 12,5 213,6/3
2005 20 80,0/4 _ _ 3 0,4 480 96,0 68 9,0/4 14 4,5/3 19 6,9 196,8/11
2006 8 15,0/1 1,5 5,0 3 0,3 316 70,5/1 73 40,0 12 2,5 11 6,2 139,5/2
2007 7 40,3/1 - - 6 1,1/1 10 3,0/1 7 5.0 49,4/3
2008 16 38/1 - - 5 2.9 4 1.9 5 2.9 45,7/3
Total 126 385/16 14,5 375,0 41 88,3/11 3781 763,1/17 1816 282,0/18 124 44,7/17 136 142.6 2080,7/79
Hail
Total
Loss
(mln.
GEL/Man)
No Data
Hazardous events intensity and caused damages and human losses during the 1995-2008 years over the territory of Georgia
Y
e
a
r
Flood Drough Strong wind Landslide Mudflow Avalanche

-


1
9
8
6
1
9
8
7
1
9
8
8
1
9
8
9
1
9
9
0
1
9
9
1
1
9
9
2
1
9
9
3
1
9
9
4
1
9
9
5
1
9
9
6
1
9
9
7
1
9
9
8
1
9
9
9
2
0
0
0
2
0
0
1
2
0
0
2
2
0
0
3
2
0
0
4
2
0
0
5
2
0
0
6
2
0
0
7
2
0
0
8
1. . _ . 220/260 231
2. . _ . 320/500 349 504 409 404 434 419 350 420 437 474
3. . _ . 300/450 354 466 400 462 367 420 400 418 511 395 428 422 390 425 411 376 389 416 352
4. . _ . 250/380 434 331 375 366 320 373 404 270 374 359 369 350 350 310 300 350 370 376 476 297 355 380
5. . _ . 650/740 950 785 822 785 705 815 915 920 865 815 660 690 665 836 665 741 730 720 720
6. . _ . 300/350 300 300
7. . _ . 300/430 395 560 320 364 377 312 327 314
8. . _ . 160/190 336 381 266 375 339 339 363 386 330 340 236 346 290 326 236 248 240 344 168 164
9. . _ . 500/560 531 558 531 571 571 552 581 501 516 501
10. . _ 340/370 351 490
11. . _ . 230/420 275 359 374 304 379 284 301 389 254 264 274 369 339 364 281 361 394 245 249 419 239 318
12. . _ 300/370 405 440 304
13. . () _ . 180/440 254 211 239 184 182 238 250 196 186 236 290 262 180 220 192
14. . () _ . 350/500 511 386 406 426 350 490 368 376 354 406 351 421 354
15. . () _ . 350/600 512 481 365 472 384 448 600 364 382 350 465 370 565 370 445
16. . _ . 260 280 274
17. . _ 140/220 171 218 227 199 250 188 218 160 230 178 148 259
18. . _ 210 210 254 216 222 233
19. . _ . 240 244 245 268 298 282 205 275
20. . _ .
21. . _ . 450/580 530

APPENDIX 2: ()
APPENDIX 3
Nr. Task Manpower need Costs Equipment and Costs
months (USD) software needs (USD)
1 Knowledge Base (Key Element 1)
1.1 Continuing of small (gross) scale mapping of hazards zones: 2 GIS experts GIS maps (1:50.000) 5000
avalanches, landslides, flood plain, mud flow * 3 years = Software (Arc lic.)*3 9000
man-months 66 33000 Hardware (3 PCs) 3000
1.2 Assessment of hazard and risk levels of lanslides and mud flows 4 geologists / GIS maps (1:10.000) 6500
geophysicists Software (Arc.lic) 3000
* 3 years = Hardware (3 PCs) 3000
man-months 132 66000
1.3 Saving and structuring of hydromet observations, measurements
and other data from the past
(only for disaster predictions) 3 technicians GIS maps 4500
1 meteorolog Software
1 hydrologist (Geodatbase) 5000
1 (IT) data manager Hardware (PCs,
*3 years = scaner, etc.) 4500
man-months 165 82500
1.4 Flood mapping of 4 priority flood plains 2 hydrology GIS maps 4700
(Mtkvari River Basin, districts: Gori, Kaspi and Tbilisi) / hydraulic experts Arc software 3000
(Alezani River Basin: left bank [Great Caucasus) 2 GIS specilists Hydraulic soft 3200
(Rioni River Basin: downstream of Kutaisi) *2, 5 years = Hardware (3 PCs) 3000
(Inguri River Basin)
man-months 12 6000
2 Monitoring and Prediction Serevice
(Key Element 2)
2.1 Monitoring
2.1.1 River Monitoring Telemetr. riv. stations
(full programme):
Mtkvari R.B. 4
Alezani R.B. 3
Rioni R.B. 4
Inguri R.B. 4
Total: 15 180000
2.1.2 Precipiation monitoring Telemetr. rain gauges:
Mtkvari R.B. 5
Alezani R.B. 5
Rioni R.B. 7
Inguri R.B. 5
Total: 22 77000
2.1.3 Field works (covers field works in knowledge base development,
operational monitoring, hydrometric surveys etc.)
Snow survey 4 specialists Snow measuring
man-months/year 3 months/year field equipments
Flow measurements 4 specialists Current meters (5) 5000
man-months per year 4 months/year Doppler (1) 21000
Flood plain surveys 3 specialists GPS
man-months/year 5 months/year 1 advanced 1200
3 general 900
River bed surveys 3 specialists Acoustic sonar (1) 1400
man-months/year 4 months/year GPS 450
Mud flow, landslide, avalanche surveys 3 specialists
man-months per year 4 months/year
67 33500
2.1.4 Assessment of geodinamical state of the relief over Georgia 10 speicalists field equipmet 30000
2 years GIS maps 2400
man-moths 110 55000
2.1.5 Baseline study of mudflow occurance 3 specilists
(analysis of mudflow events) 3 years
man-months 33 16500 geostatistical softw 3700
hardware 4800
2.2 Prediction
2.2.1 Spring flood prediction
Downscaling the grid of the regional weather forecasts 1 IT specilaists GIS sowtware 2300
from 14*14 km to 7*7 km 1 meteorologists hardware (2 PCs) 2000
2* 0,75 years =
man-months 16 8000
Short-term (flash flood) prediction 1 IT/modeller GIS license 2700
adopting and operational test of forecasting model for Rioni River 1 hydrologist hardware 3000
1 meteorologists
3*1 year =
man-months 33 16500
Short-term (flash flood) prediction 1 IT/modeller
adopting and operational test of forecasting model for Rivers 2 hydrologist
Inguri, Kura and Alazani 1 meteorologists
4*2 years =
man-months 88 44000
3 Dissemination and Communication (Key Element 3)
3.1 Integrated Database Management at NEA
3.1.1 System design and data inventory 1 IT specialists server 7500
2 hydrologists software 3300
2 meteorologists for data management
hardware 8000
5*8 months
man-months 48 24000
3.1.2 Implementaion of the system 2 IT specialists
2*1 year
man-months 22 11000
3.2.1 Integrated Communication System for Early Warning of
Natural Disasters
3.2.2 System design 2 IT specialists
Specialists:
2 from NEA software license 12000
2 from MIA hardware needs
6 from different at different agencies 25000
agencies
12*4 months 24000
Implementaion of the system 2 IT specialists
2*1 year
man-months 22 11000
4 Response Capability
4.1 National Response Strategy and Planning
4.1.1 Desighn annual national planning template document
1 international
consultant
1 public policy
specialist
1 legal specialist
1 risk assesment
specialist
4*6 month
man-months 24 21,000
4.1.2. Desighn annual sectoral planning templates:
1 international
consultant
MIA/EMD, MEPNR, MRDI, MOH, MED, MOA, MFA, MOD 1 public policy spec.
1 legal specialist
1 risk assesment
specialist
4*12 month
man-months 48 72,000
4.2 Regional/local Response Strategy and Planning
4.2.1 Desighn annual local planning template document
1 intermational
consultant
1 public policy
specialist
1 legal specialist
1 risk assesment
specialist
4*12 month
man-months 48 72,000
4.3 Public Awarness and Education
4.3.1 Desighn National Public Awarness Staretgy Plan for Dissater Risk Reduction and Natural Hazards Preparadness
1 International
consultant
1 public policy
speciailst
1 legal specialist
1 risk assesment
specialist
man-months 4*12 month
48 72,000
4.3.2 Desighn National Curricula on the Natural Hazards Preparadness to be Introduced at Primary Schools Nationwide
1 International
consultant
1 education policy
speciailst
1 natural sciences
specialist
3*24 month
man-months 72 72,000
Nr. Task Manpower need Competence development need Notes
1 Knowledge Base (Key Element 1)
1.1 Continuing of small (gross) scale mapping of hazards zones: 1 specialist International training in hazard and
avalanches, landslides, flood plain, mud flow 2 GIS experts risk mapping
man-months * 3 years =
99
1.2 Assessment of hazard and risk levels of lanslides and mud flows 4 geologists /
The expereinced specialists shall
train
geophysicists young experts
* 3 years =
man-months 132
1.3 Saving and structuring of hydromet observations, measurements
and other data from the past
3 technicians No need for special training
1 meteorolog
1 hydrologist
1 (IT) data manager
*3 years =
man-months 165
1.4 Flood mapping of 4 priority flood plains 2 hydrology
(Mtkvari River Basin, districts: Gori, Kaspi and Tbilisi) / hydraulic experts
International ltraining in flood
mapping training
(Alezani River Basin: left bank [Great Caucasus) 2 GIS specilists and supervision
(Rioni River Basin: downstream of Kutaisi) *2, 5 years =
(Inguri River Basin) 110
2 Monitoring and Prediction Serevice
(Key Element 2)
2.1 Monitoring
Instalation and operation of the telemetric network 3 technicians
Training in instalation and
operation of
the equipments
2.1.3 Field works (covers field works in knowledge base development,
operational monitoring, hydrometric surveys etc.)
Snow survey 4 specialists Training for using new equipment
man-months/year 3 months/year
Flow measurements 4 specialists Training for using new equipment
man-months per year 4 months/year
Flood plain surveys 3 specialists Training for using advanced GPS
man-months/year 5 months/year
River bed surveys 3 specialists Training for using new equipment
man-months/year 4 months/year
Mud flow, landslide, avalanche surveys 3 specialists Taining for using new equipment
man-months per year 4 months/year
2.2 Prediction
2.2.1 Spring flood prediction
Downscaling the grid of the regional weather forecasts 1 IT specilaists
International training needed in
dowscaling
from 14*14 km to 7*7 km 1 meteorologists
of regional weather forecast
outputs.
2* 0,75 years =
man-months 16
Short-term (flash flood) prediction 1 IT/modeller
International training needed in
applying
adopting and operational test of forecasting model for Rioni River 1 hydrologist
advanced flash flood forecasting
models
1 meteorologists
3*1 year =
man-months 33
Short-term (flash flood) prediction 1 IT/modeller
adopting and operational test of forecasting model for Rivers 2 hydrologist
Inguri, Kura and Alazani 1 meteorologists
4*2 years =
man-months 88
3 Dissemination and Communication (Key Element 3)
3.1 Integrated Database Management at NEA
International training needed in
database management
3.1.1 System design and data inventory 1 IT specialists
2 hydrologists
2 meteorologists
5*8 months
man-months 48
3.1.2 Implementaion of the system 2 IT specialists
2*1 year
man-months 22
3.2.1 Integrated Communication System for Early Warning of
International training and
supervision
Natural Disasters
needed in setting up the integrated
system
3.2.2 System design 2 IT specialists
Specialists:
2 from NEA
2 from MIA
6 from different
agencies
12*4 months
Implementaion of the system 2 IT specialists
2*1 year
man-months 22
4 Response Capability
4.1 National Response Strategy and Planning
4.1.1 Desighn annual national planning template document 1 international consultant International supervision needed
1 public policy specialist
1 legal specialist
1 risk assesment specialist
4*6 month
man-months 24
4.1.2. Desighn annual sectoral planning templates: 1 international consultant International supervision needed
MIA/EMD, MEPNR, MRDI, MOH, MED, MOA, MFA, MOD 1 public policy specialist
1 legal specialist
1 risk assesment specialist
4*12 month
man-months 48
4.2 Regional/local Response Strategy and Planning
4.2.1 Desighn annual local planning template document 1 intermational consultant International supervision needed
1 public policy specialist
1 legal specialist
1 risk assesment specialist
4*12 month
man-months 48
4.3 Public Awarness and Education
4.3.1
Desighn National Public Awarness Staretgy Plan for Dissater Risk
Reduction and Natural Hazards Preparadness 1 International consultant International supervision needed
1 public policy speciailst
1 legal specialist
1 risk assesment specialist
man-months 4*12 month
48
4.3.2
Desighn National Curricula on the Natural Hazards Preparadness to be
Introduced at Primary Schools Nationwide 1 International consultant International supervision needed
1 educ. policy speciailst
1 natural sciences specialist
3*24 month
man-months 72