30

M.K. Shrestha, S. Chaudhary, R.K. Maskey & G. Rajkarnikar December 2010

Comparison of the Anomaly of Hydrological Analysis

Tools used in Nepal
M. K. Shrestha, S. Chaudhary1, R.K. Maskey2 & G. Rajkarnikar3
Email:manojksth@gmail.com
1
B. Tech. Environmental Engineering, Kathmandu University.
2
Profesor for Hydraulics and Hydropower Engineering School of Engineering Kathmandu University.
3
Senior Divisional Hydrologist Water and Energy Commusion Secretariat Kathmandu.

Abstract

The diverse physiographic and hydrological regime has booned Nepal with high potential of water resource
projects. But these projects depend highly on hydrological data for which hydrometeorological stations need
to be established. Their establishment, because of difficult topographic feature of Nepal and the high cost of
installation, operation and maintenance, becomes feasible only for big projects. And thus hydrological data are
needed to be predicted for small scale projects. These predictions for a given ungauged river system at particular
location are facilitated by empirical methods such as WECS/DHM (Water and Energy Commission Secretariat/
Department of Hydrology and Meteorology) and MIP (Medium Irrigation Project) which have been used since its
development without upgrading and checking its reliability. The objective of this study is to compare the anomaly
of flow predicted by WECS/DHM and MIP method and determine reliability of minimum flow predicted by the
methods. The anomaly of the method is presented by comparing the mean DHM monthly flow with the mean of
predicted monthly flow for each of the seven rivers from Central and Western Development Region. Both methods
give variable deviation for different periods of time for all the rivers. MIP method gives reliable prediction only
if the discharge measurement is done during the dry period (November–April). The reliability of these methods
is checked for estimating minimum flow by calculating the percentage of time discharge (minimum predicted
flow) will be exceeded. WECS/DHM has given higher reliability for the minimum flow than MIP method. MIP
(D) method has given good approximation to the DHM dry period flow than MIP (A). While analyzing the low
flow, lower predicted value given by either of these methods when used gives good approximation. In order to
improve the accuracy of prediction there is requirement of modification of these Hydrological Analytical tools.

Keywords: hydrological data; WECS/DHM method; MIP method; ungauged river;

MIP (D); MIP (A)

1. INTRODUCTION Nepal lack the sufficient hydro meteorological

data because of which vast amount of water
Nepal is rich in water resources and utilization resources has remained unexploited at present.
of water resources through projects such as In order to evaluate, compare and ultimately
hydropower or irrigation depends highly upon develop potential projects hydrological variables
the hydrological data. Most of the locations of of temporal flow condition should be regularly

Journal of Hydrology and Meteorology, Vol. 7, No. 1 SOHAM-Nepal

December 2010 M.K. Shrestha, S. Chaudhary, R.K. Maskey & G. Rajkarnikar 31

monitored. Developing a complete network regions. No regionalizations were done for

of hydro-meteorological stations for country either low flows or flood flows. It can only
like Nepal is expensive. Moreover problems give the mean monthly flows.
such as lack of efficient manpower, allocation
of less resource in collection and processing 2. Rationale of Study
of hydrologic data and difficult topographic
feature are the hindrances for its establishment. These two methods are in use for the estimation of
Major water resource projects establish their the flow conditions for various medium-small scale
own gauging stations good estimation of flow water resource projects at the ungauged locations.
conditions regarding the water availability and When WECS/DHM method was developed, the
risk minimization relating to the floods. But this country had only 54 hydrological stations which
is not possible for small scale projects. So it led to were insufficient to completely represent the
the need of methods that could estimate the flow country in terms of hydrological regime. So it has
conditions in the ungauged catchment. Several made several recommendations of updating the
methods and models have been developed for method every five years, developing hydrometric
regionalization of hydrologic behavior in Nepal. network review and plans, making extensive
The most frequently used are: precipitation study including the Himalayan
region and Siwaliks etc.
1. WECS/DHM method
WECS/DHM has inferred the MIP method as
It is a modification of WECS approach of less comprehensive study. This design manual
1982 and has been developed jointly by is intended for guidance only and consequently
WECS and DHM in cooperation with WMO it is not as complete, nor as rigorous, as the
(World Meteorological Organization), WECS (1982) study (WECS/DHM, 1990). The
WERDP (Water and Energy Resource hydrograph developed should be checked or
Development Project, until 1989) and updated so that the prediction made on the basis
WISP (WECS/NEA Institutional Support of the hydrograph is accurate and the discharge
Programme) in 1990. It treats the entire predicted could be used.
country as a single hydrological region. The
regionalization was done for low flows, long But these recommendations have not been
term flows and flood flows. realized and the method is in use since they are
developed. Since there is no other choice for the
2. MIP (Medium Irrigation ungauged locations; these methods are widely
Project) method relied upon by the water resource projects. As
these tools didn’t meet the recommendations, it is
It was developed in 1982 by Sir M. MacDonald not wise to fully depend upon their predictions.
and Partners Limited in association with So this study is to check the anomaly and
Hunting Technical Services Limited in reliability so that the risk associated with their
which Nepal is divided into 7 hydrological predictions will be timely avoided.

Journal of Hydrology and Meteorology, Vol. 7, No. 1 SOHAM-Nepal

32 M.K. Shrestha, S. Chaudhary, R.K. Maskey & G. Rajkarnikar December 2010

3. Objective of Study the entire country as a single entity. The two major
factors affecting the hydrological characteristics
As their reliability can be questioned, we have the of river is location of catchment area and rainfall
following objectives: in the catchment. Monsoon rainfall contributes
to flood runoff which is generally below 3000m
a) To compare anomaly of predicted where as the river above 5000m is usually covered
discharge data by WECS/DHM method with snow mainly contributing to long term flow.
and MIP method with gauging station The MIP method is based upon measurement
data of DHM taken on an intermittent basis. The measurement
b) To determine the percentage of time, of lowest discharge usually April is used to predict
discharge will be exceeded for the lowest the mean monthly discharge of a particular location
discharge predicted by the methods using a Unit Hydrograph (l/s per sq. Km) which
was used to develop Non-dimensional hydrograph
4. Literature Review for seven regions.

Many regional methods for estimating the flow 5. METHODOLOGY

conditions at ungauged sites have been developed
in different parts of the world (e.g. Singh (1971); Seven rivers whose hydrological data has
Quimpo et al. (1983); Finnessey and Vogel (1990)). been recorded by DHM were selected whose
These regionalization methods assume that catchment lies in the MIP regions- either 1 or 3
within a given region, catchments having similar given in Table 1. We selected those rivers whose
climate, geology, topography, vegetation and soils catchment ranges within 1000 sq. km and lying
display similar hydrological responses (Smakhtin, in Western and Central Development Region
2001). So, homogenous hydrologic regions are of Nepal. Two rivers from MIP Region 1 and
selected for this type of study. Long term and five rivers from MIP Region 3 were selected.
low flow hydrology is particularly dependent The hydrological data from the year 1976-
upon topography and geology. Hence, different 2006 were taken into account with five years
physiographic regions with uniform topographic of interval except for Langtang (1993 onward)
and geologic characteristics are frequently and Melamchi (1990 onward). 15 meteorological
selected as homogenous hydrologic regions. In stations were selected to calculate the monsoon
Nepal, this is not possible because the drainage wetness index. For each of the river basin at least
basins run approximately north- south while 3 representative meteorological stations were
the physiographic regions run approximately selected with exception to the Langtang River(2).
east- west. Hydrologic response of the basins as The Catchment area of the respective river was
represented by the available stream gauge network determined using the Topographical maps of
is a result of the physical response of several of the scale 1:25,000 prepared by Survey Department
physiographic regions contained within the basins. of Government of Nepal in co-operation with
As a result of this problem, it was decided to treat government of Finland.

Journal of Hydrology and Meteorology, Vol. 7, No. 1 SOHAM-Nepal

December 2010 M.K. Shrestha, S. Chaudhary, R.K. Maskey & G. Rajkarnikar 33

Table 1: Seven rivers of Central and Western development region

Mip Station Drainge Area
Name of River Location Latitude Longitude Elevation (M)
Region Number (Km2)
1 Melamchi Helambu 627.5 28º 02΄ N 85º 32΄ E 2134 84
1 Langtang Kenjying 446.1 28º 13΄ N 85º 37΄ E 3920  333
3 Khimti Rasnalu 650 27º 34΄ N 86º 11΄ E 1120 313
3 Solu Khola Salme 668.5 27º 30΄ N 86º 34΄E 1800 246
3 Madi Khola Shisaghat 438 28º 06΄ N 84º 14΄ E 457 858
3 Chepe Khola Gharm besi 440 28º 03΄ N 84º 29΄ E 442 304
3 Tadi Belkot 448 27º 51΄ N 85º 08΄ E 475 653

The flow was calculated by using both methods. methods in predicting the low flow. This curve
As MIP gives the flow calculation of mean is obtained by plotting the discharge against the
monthly flows, we have considered the calculation percentage of time the minimum predicted flow
of long term flows only for WECS/DHM method will be exceeded. The discharge taken is the DHM
for similar comparison. The DHM Stream flow minimum flow of each year. The percentage of
Summary data of corresponding river is taken as time minimum flow predicted by both methods
reference for comparison. will exceed is then calculated from Flow Duration
Curve of minimum flow.
To determine the reliability among the methods,
comparison of WECS/DHM Method and MIP 6. RESULTS AND DISCUSSION
Method (using April flow and the driest monthly
flow) against the DHM summary flow data was The anomaly of methods from the DHM data was
made. April is considered to be the driest month done on the basis of mean monthly discharge.
and so April flow is recommended by the literature This has given variable result which is shown
for calculation of flows from MIP method. But the in Figures 1- 7. WECS/DHM gives negative
DHM summary flow data of seven rivers shows that deviation for all the months for Melamchi, Khimti,
this consideration does not hold true for all the years Solu, Madi and Chepe. Comparison against
and so the attempt is made in this project to predict Langtang gives the positive deviation for June –
the flow using MIP method by considering both December (during monsoon and post monsoon
driest and April flow ,designated as MIP(D) and period) which indicates that the modification of
MIP(A), and compare the result of both. Anomaly the tool is required for the snow fed rivers. But
of WECS/DHM and MIP method is calculated in comparison against Tadi gives positive deviation
terms of percentage difference from DHM values. for March – May (during pre-monsoon) which
The percentage difference is calculated on the basis shows that the overall generalization cannot
of mean discharges. Anomaly shows percentage be made for all the rivers of Nepal. MIP (D)
from which the predicted values are positively or gives both positive and negative deviation for
negatively deviated for the respective months from both MIP Region I and III. For Region I, MIP
mean DHM values. (D) gives negative deviation for April and June
for Langtang and for Melamchi gives negative
Flow Duration Curve of minimum flow was deviation for March – November. For Region III,
plotted to determine the reliability of these the negative deviation persists between March-

Journal of Hydrology and Meteorology, Vol. 7, No. 1 SOHAM-Nepal

34 M.K. Shrestha, S. Chaudhary, R.K. Maskey & G. Rajkarnikar December 2010

July for all rivers except Tadi. It gives negative and Solu and for other rivers it gives negative
deviation for February and even for April- deviation for the period between June – July
November (pre monsoon and monsoon period). (Khimti), June (Madi), April (Langtang), April
MIP (A) gives positive deviation for Melamchi – July (Chepe), and June –August (Tadi).

250
225
200
175
150
125
100 WECS/DHM
75
MIP(D)
50
% 25 MIP(A)
DIFFERENC
E 0
-25
-50
-75
-100
-125
JanF eb Mar Apr MayJ uneJ uly Aug SeptO ct NovD ec

WECS/DHM -70.8- 71.6- 74.6- 82.2- 89.0 -80.0- 77.9 -73.1- 74.3- 71.0 -68.6- 58.6
MIP(D) 45.71 24.51 -9.23- 50.1- 38.1- 45.6- 43.9- 3.55 -21.9- 1.66 -0.05 53.55
MIP(A) 213.7 168.0 95.44 7.27 33.24 17.05 20.75 107.6 68.02 111.7 115.1 230.6

Figure 1: Anomaly of WECS/DHM and MIP – Melamchi

350
325
300
275
250
225
200
175 WECS/DHM
150
125 MIP(D)
100 MIP(A)
75
50
%
25
DIFFERENCE
0
-25
-50
-75
JanF eb Mar Apr MayJ uneJ uly Aug Sept OctN ov Dec
WECS/DHM -36.5- 43.1- 36.5- 40.3 -52.02 2.49 17.92 43.41 39.703 9.31 81.413 8.68
MIP(D) 41.04 11.120 .09- 28.5 12.78 -6.64 79.15 213.2 168.0 185.5 99.05 72.56
MIP(A) 88.35 48.40 33.67 -4.545 0.62 24.69 139.2 318.3 257.9 281.3 165.8 130.4

Figure 2: Anomaly of WECS/DHM and MIP – Langtang

Journal of Hydrology and Meteorology, Vol. 7, No. 1 SOHAM-Nepal

December 2010 M.K. Shrestha, S. Chaudhary, R.K. Maskey & G. Rajkarnikar 35

225

200

175

150

125

100

75 WECS/DHM
50
MIP (D)
25
MIP(A)
%
0
DIFFERENCE
-25

-50

-75

-100
JanF eb MarA pr MayJ uneJ ulyA ug Sept OctN ov Dec
WECS/DHM -37.17 -37.48 -39.32 -38.76 -52.06 -63.73 -50.39 -33.63 -31.56 -23.89 -36.41 -33.75
MIP(D) 82.83 48.23 16.12 -19.63 -14.30 -66.94 -36.75 30.65 46.83 86.34 73.25 108.52
MIP(A) 175.23 123.14 74.80 20.98 29.01 -50.23 -4.78 96.68 121.04 180.52 160.81 213.90

Figure 3: Anomaly of WECS/DHM and MIP – Khimti

200
175
150
125
100
75 WECS/DHM
50 MIP(D)
% 25 MIP(A)
DIFFERENC
E 0
-25
-50
-75 A r A
-100
JanF eb Mar p MayJ une July ug SeptO ct NovD ec
WECS/DHM -60.1- 59.9 -59.9- 59.6 -65.6- 37.7- 42.5- 49.0- 34.7- 45.9- 58.8- 57.5
MIP(D) 76.38 44.42 16.85 -18.2- 4.11 -14.01 1.32 51.96 111.7 100.0 67.85 100.0
MIP(A) 129.2 87.72 51.89 6.29 24.64 11.75 44.70 97.51 175.1 160.0 118.1 160.0

Figure 4: Anomaly of WECS/DHM and MIP – Solu Khola

200
175
150
125
100
75 WECS/DHM
50 MIP(D)
% 25 MIP(A)
DIFFERENC
E 0
-25
-50
-75
-100
JanF eb Mar Apr MayJ uneJ uly Aug SeptO ct NovD ec

WECS/DHM -37.1- 40.6 -42.2- 47.1 -54.7- 58.3 -40.2- 27.3 -20.9- 17.9 -30.9- 35.6
MIP(D) 57.84 21.44- 6.33 -44.0- 37.7- 70.3- 40.9 11.93 32.65 59.104 6.68 60.42
MIP(A) 181.8 116.8 67.28 0.00 11.23 -47.05 .43 99.90 136.8 184.1 161.9 186.4

Figure 5: Anomaly of WECS/DHM and MIP– Madi Khola

Journal of Hydrology and Meteorology, Vol. 7, No. 1 SOHAM-Nepal

36 M.K. Shrestha, S. Chaudhary, R.K. Maskey & G. Rajkarnikar December 2010

150
125
100
75
WECS/DHM
50
MIP(D)
25
% MIP(A)
0
DIFFERENCE
-25
-50
-75
Jan FebM ar Apr May June July Aug Sept OctN ov Dec

WECS/DHM -32.7 -34.3- 34.8- 32.1- 38.0- 42.4- 27.8- 25.1- 25.0 -20.2- 29.1- 28.0
MIP(D) 57.41 25.27 0.29 -28.2- 10.7- 59.8- 29.4 12.94 22.51 49.88 44.37 70.11
MIP(A) 113.7 70.12 36.20 -2.55 21.26 -45.4- 4.12 53.38 66.37 103.5 96.05 131.0

Figure 6: Anomaly of WECS/DHM and MIP - Chepe Khola

80
60
40
20 WECS/DHM
%
0 MIP(D)
DIFFERENCE
-20 MIP(A)

-40
-60
-80
Jan FebM ar Apr MayJ une July Aug Sept Oct NovD ec

WECS/DHM -14.55 -5.41 30.76 48.80 15.95 -25.90- 7.23 -19.55 -14.74 -16.43 -25.25 -16.19
MIP(D) 8.21 -2.30 7.52 -19.01 -16.91- 72.79 -52.60 -36.07 -26.23 -16.68 -18.25 7.18
MIP(A) 68.64 52.26 67.57 26.22 29.49 -57.59 -26.13 -0.37 14.96 29.84 27.41 67.04

Figure 7:Anomaly of WECS/DHM and MIP - Tadi Khola

Journal of Hydrology and Meteorology, Vol. 7, No. 1 SOHAM-Nepal

December 2010 M.K. Shrestha, S. Chaudhary, R.K. Maskey & G. Rajkarnikar 37

7. Result of Flow Duration if the driest flow could be measured; otherwise this
Curve of minimum flow reliability will be reduced as given by MIP (A). As
the non dimensional hydrograph ordinate remains
The reliability of the lowest predicted flow with the same, the greater the flow taken more is the increase
minimum flow of each year for respective river in the ordinate of the predicted hydrograph. April
from 1976 – 2006 was calculated by determining is considered to be driest but from the hydrological
the percentage of time the discharge (minimum data of the seven rivers selected the average driest
predicted flow) will be exceeded for both the month is March except for Chepe which is April.
methods which is given in Table 2. WECS/DHM In the case, when April flow is not the driest, the
gives the flow availability of up to 100% during April flow if taken gives higher predicted value
dry period for Solu Khola. The percentage of than the driest flow taken i.e., MIP (A) > MIP (D)
reliability for low flow given by WECS/DHM is . So, continuous measurement of the flow between
higher than MIP except for Tadi. The data quality November-April is required to obtain the driest
of Tadi is ranked to be good by DHM and so this flow. But in the case when this measurement is
result cannot be excluded. For MIP (D) the flow not possible it is suggested that the flow should
availability varies between 24 – 86 % of time be estimated from both the methods and the flow
during dry period. But this reliability is valid only should be so chosen that has lower values.

Table 2: Reliability of minimum flow

Predicted minimum flow (m3/s) % of Time Q will be exceeded

for minimum flow
Range of minimum yearly
River WECS/ WECS/
flow of DHM (m3/s) MIP MIP
MIP (A) MIP (A)
(D) (D)
DHM DHM
Melamchi 0.47 (2000) – 3.20 (2005) 0.69 1.94 4.17 88 59 *
Khimti 1.49 (2001) – 7.26 (1990) 3.04 4.22 6.35 74 24 5
Solu 4.67 (2001) – 6.48 (1996) 2.4 5.07 6.59 100 78 *
Madi 6.34 (1993) – 17.9 (1978) 8.28 9.74 17.39 91 86 6
Chepe 1.37 (2001) – 4.9 (1990) 3 3.34 4.54 66 45 10
Tadi 0.41 (2000) – 6.47 (1976) 6.32 3.76 5.87 5 38 7

* - Predicted lowest flow exceeds the range of is higher for WECS/DHM than MIP. Among
minimum yearly flow of DHM three methods, the reliability of minimum flow
predicted by MIP (A) is worse. MIP (D) gives
Q- Predicted minimum flow good approximation than MIP (A). This higher
The percentage of time Q will be exceeded for reliability is obtained only if the driest flow is
minimum flow of each river can be inferred from taken in account.
figure of Flow duration curve of minimum flow for
respective rivers. The Table doesn’t include the data for Langtang
because of unavailability of data of minimum
For all the rivers except Tadi the percentage annual flow.
of time Q will be exceeded for minimum flow

Journal of Hydrology and Meteorology, Vol. 7, No. 1 SOHAM-Nepal

38 M.K. Shrestha, S. Chaudhary, R.K. Maskey & G. Rajkarnikar December 2010

8. CONCLUSION Department of Hydrology and Meteorology,

Mr. Shyam Sundar Khadka and Miss Silu
The prediction made by these methods gives Bhochibhoya, Research Assistant, Centre for
variable result to conclude that the individual Excellence in Production and Transportation
methods are not reliable. However the flow of Electrical Energy (CEPTE/KU), Er. Ishwor
obtained by comparing the predicted flows using Onta and Er. B.P. Shah of East Consult Pvt. Ltd.,
both methods could give good approximation but Basundhara for their help to find reports of Medium
not to the snow fed rivers. But this generates the Irrigation Project, and Mr. Suresh Maskey,
complexity of using both methods and comparing Department of Hydrology and Meteorology for
them and finding one result, which clearly shows their help during our project study.
the incapability of the single method.
References
Further MIP depends upon the measured flow.
The river flow is very variable even in the same DHM, 2008. Stream flow Summary (1962-2006),
month for different year so it is difficult to predict Ministry of Environment Science and
for future by MIP and has the probability that the Technology, Department of Hydrology and
water resource project does not sustain long. Also, Meteorology, October 2008, pp 177-357
the result it gives can have good approximation
only if the driest flow could be accounted. As the DOI, 1990. Design Manuals for Irrigation Projects
driest flow to be measured could not be predicted in Nepal, 1990. Planning and Design
it requires continuous flow measurement during Strengthening Project (PDSP), Ministry of
the dry periods. To have better future prediction Water Resources, Department of Irrigation.
sufficient data is required, which would not be United Nations Development Programme,
available for the ungauged catchment. pp 35-36.

As the number of hydrological station has increased ESAP, 2001. Guidelines Verification of Power
up to 99 from 54 in 1990, it shows the prospect Output from Micro-Hydro Plants, 2001.
modification and regular update of these methods Interim Rural Energy Fund, Alternative
in each five year interval, as recommended by Energy Promotion Centre, Energy Sector
WECS/DHM. This recommendation shall hold Assistance Programme, Kathmandu,
true for MIP as well. This will surely increase September, 2001. pp 11-41.
the reliability of individual methods and thus
help to make the water resource project well MIP, 1982. Macdonald and Partners Ltd. Medium
sustainable for years. Irrigation Project, Design Mannual,
Hydrology. July, 1982, pp 21-26.
9. ACKNOWLEDGEMENT
Quimpo, R.G. Alejendrino, A.A. and McNally,
The author would like to thank Dr. Rijan Bhakta T.A., 1983. Regionalized flow duration for
Kayastha, Assistant Professor, Department of Philippines. In: J. Water resource planning
Environmental Science and Engineering, Dr. and Management, ASCE, pp 320-330.
Keshav Prasad Sharma, Hydrology Division,

Journal of Hydrology and Meteorology, Vol. 7, No. 1 SOHAM-Nepal

December 2010 M.K. Shrestha, S. Chaudhary, R.K. Maskey & G. Rajkarnikar 39

Shrestha, H. M., 1985: Water power potential in Energy Commission Secretariat and
Nepal: Nature’s paradise, White Lotus Co. Department of Hydrology and Meteorology,
Ltd., Bangkok, pp 32-35. Kathmandu, Nepal, July 1982.

Smakhtin, V.U., 2001. Low flow Hydrology: a WECS, 1990. Methodology for estimating
review. Journal of Hydrology. pp 147-186. hydrological characteristics of ungauged
locations in Nepal, His majesty’s Government
of Nepal, Ministry of Water Resources
WECS, 1982. Hydrological Studies of Nepal. Water and Energy Commission Secretariat
Volume 1- Report, Voliume 2, 3-Appendices, and Department of Hydrology and
His majesty’s Government of Nepal, Meteorology, Kathmandu, Nepal, July 1990.
Ministry of Water Resources Water and

Journal of Hydrology and Meteorology, Vol. 7, No. 1 SOHAM-Nepal