wd = getwd()

setwd("~/git/M9/")

slipDat <<- read.table("Cascadia-lockingrate-hdr.txt", header=TRUE)

names(slipDat) <<- c("lon", "lat", "slip", "Depth", "slipErr")

slipDat$Depth <<- slipDat$Depth*10^3 # convert from km to m

attach(slipDat)

lonRange <<- range(lon)

latRange <<- range(lat)

# fault geometry data (convert from km to m and to longitude format consistent with other data)

faultGeom <<- read.csv("CSZe01.csv")

faultGeom$longitude <<- faultGeom$longitude - 360

kmCols <<- c(4, 6, 7)

faultGeom[,kmCols] <<- faultGeom[,kmCols] * 10^3

csz <<- divideFault(faultGeom)

# subsidence data

load("DR1.RData", envir=globalenv())

attach(dr1)

# get unique CSZ earthquake event names

events <<- as.character(event)

uniqueEvents <<- unique(events)

sortI = rev(c(1, 12, 11, 14, 2, 15, 3, 4, 18, 5, 13, 6, 19, 7, 20, 10, 21, 8, 9, 16, 17)) # T1 is most recent, so reverse (T1 is last)

uniqueEvents <<- uniqueEvents[sortI]

# subset uniqueEvents so it only contains major events

eventSubset = c(1, 2, 4, 6, 8, 10, 12, 15, 17, 19, 20, 21)

majorEvents = uniqueEvents[eventSubset]

# remove all other events from dr1

nonMajorEvents = setdiff(uniqueEvents, majorEvents)

for(i in 1:length(nonMajorEvents)) {

nonEventInds = events != nonMajorEvents[i]

dr1 <<- dr1[nonEventInds,]

}

uniqueEvents <<- majorEvents

# sort dr1 by event

dr1$event <<- factor(dr1$event, levels=uniqueEvents)

inds = order(dr1$event)

dr1 <<- dr1[inds,]

events <<- as.character(dr1$event)

#set working directory back to what it was before calling this function

setwd(wd)

invisible(list(dr1=dr1, slipDat=slipDat, csz=csz))

wd = getwd()

setwd("~/git/M9/")

slipDat <<- read.table("Cascadia-lockingrate-hdr.txt", header=TRUE)

names(slipDat) <<- c("lon", "lat", "slip", "Depth", "slipErr")

slipDat$Depth <<- slipDat$Depth*10^3 # convert from km to m

attach(slipDat)

lonRange <<- range(lon)

latRange <<- range(lat)

# fault geometry data (convert from km to m and to longitude format consistent with other data)

faultGeom <<- read.csv("CSZe01.csv")

faultGeom$longitude <<- faultGeom$longitude - 360

kmCols <<- c(4, 6, 7)

faultGeom[,kmCols] <<- faultGeom[,kmCols] * 10^3

csz <<- divideFault(faultGeom)

# subsidence data

load("DR1.RData", envir=globalenv())

attach(dr1)

event <<- dr1$event

# get unique CSZ earthquake event names

events <<- as.character(event)

uniqueEvents <<- unique(events)

sortI = rev(c(1, 12, 11, 14, 2, 15, 3, 4, 18, 5, 13, 6, 19, 7, 20, 10, 21, 8, 9, 16, 17)) # T1 is most recent, so reverse (T1 is last)

uniqueEvents <<- uniqueEvents[sortI]

# sort dr1 by event

dr1$event <<- factor(dr1$event, levels=uniqueEvents)

inds = order(dr1$event)

dr1 <<- dr1[inds,]

events <<- as.character(dr1$event)

# divide uncertainty by 2 because Uncertainty seems to represent 1.96 sigma

dr1$Uncertainty <<- dr1$Uncertainty/qnorm(0.975)

# subset GPS data by whether it's in the CSZ fault geometry boundaries

slipDatCSZ <<- getFaultGPSDat()

# piecewise linear spline

coastLonLat = matrix(c(-127, 50,

-124.8, 49.2,

-125.85, 49.1,

-124.3, 48.2,

-123.8, 46.3,

-124.1, 43.7,

-124.5, 42.85,

-124.1, 40.85,

-124.4, 40.5,

-124, 40), byrow = TRUE, ncol=2)

coastLon <<- coastLonLat[,1]

coastLat <<- coastLonLat[,2]

getSubPoints = function(nPoints) {

# get some points from the line in the piecewise spline.

# first get the latitudes of the points

maxLat = max(coastLat)

minLat = min(coastLat)

allCoastLats = seq(minLat, maxLat, length=nPoints)

# now find alpha (for the interpolation) for each section.

# take a weighted average in between each point

# for the given latitude, compute the longitude for the

# piecewise spline

getLon = function(lat) {

# first deal with base case

if(lat == 40)

return(coastLon[length(coastLon)])

# get index of point that is the first smaller than the given lat

last = match(TRUE, coastLat < lat)

# now get the point right before the first point smaller

first = last-1

# get coordinates of the points to interpolate

lastCoords = coastLonLat[last,]

firstCoords = coastLonLat[first,]

# interpolate (alpha and 1-alpha are weights for first and last)

alpha = (lat - lastCoords[2])/(firstCoords[2] - lastCoords[2])

# now make sure to do the interpolation

return(alpha*firstCoords[1] + (1-alpha)*lastCoords[1])

}

# Now vectorize the above function

allLon = sapply(allCoastLats, getLon)

allCoords = cbind(allLon, allCoastLats)

return(allCoords)

}

allCoastCoords <<- getSubPoints(250)

allCoastLon <<- allCoastCoords[,1]

allCoastLat <<- allCoastCoords[,2]

# read in Goldfinger 2012 age data

# ages <<- read.csv("goldfingerAgeEsts.csv", header=TRUE)

# remove all data past T13 since it won't be used to calculate interevent times for

# land-based coseismic subsidence data (which only goes up to T12)

# ages <<- ages[1:33,]

# change names of T10b and T10f for consistency with dr1 dataset

# ages[,1] <<- as.character(ages[,1])

# ages[25, 1] <<- "T10R1"

# ages[29, 1] <<- "T10R2"

# ages[,1] <<- factor(ages[,1])

# # remove events not in dr1 dataset:

# # Events in ages that aren't in dr1: T2a, T5c, T6b, T8b, T9b, T10a, T10c, T10d, T10e, T10f

# eventsToDelete = c(3, 11, 14, 19, 24, 26, 27, 28, 29)

# ages <<- ages[-eventsToDelete,]

# make sure ages variables are in correct format

# ages$age <<- as.numeric(as.character(ages$age))

# calculate interevent times

# age1 = ages[1:(nrow(ages)-1),2]

# age2 = ages[2:nrow(ages),2]

# intereventTime = c(age2 - age1, NA)

# ages$intereventTime <<- intereventTime

# majorEvents = c(1, 2, 4, 6, 8, 12, 15, 17, 20, 23, 30, 31, 33)

# majorAges <<- ages[majorEvents,]

# majorAges1 = majorAges[1:(nrow(majorAges)-1),2]

# majorAges2 = majorAges[2:nrow(majorAges),2]

# majorIntereventTime = c(majorAges2 - majorAges1, NA)

# majorAges$intereventTime <<- majorIntereventTime

# goldMwAll <<- read.csv("goldfinger2012Table8.csv", header=TRUE)

# # convert from dyne cm to Nm

# goldMwAll$seismicMoment <<- goldMwAll$seismicMoment * 10^(-7)

# dr1Events = c(1:10, 12:17, 19:20, 22, 24, 28:29)

# dr1Events = 1:31 # only keep events after and including T13

# goldMw <<- goldMwAll[dr1Events,]

#set working directory back to what it was before calling this function

setwd(wd)

invisible(list(dr1=dr1, csz=csz, slipDat=slipDat, slipDatCSZ=slipDatCSZ))

# helper function for determining if GPS data is within a specific subfault geometry

getSubfaultGPSDat = function(i) {

row = faultGeom[i,]

geom = calcGeom(row)

corners = geom$corners[,1:2]

in.poly(cbind(dat$lon, dat$lat), corners)

}

# construct logical matrix, where each column is the result of getSubfaultGPSDat(j)

# Hence, row represents data index, column represents subfault index. If a data

# observation is in any subfault, it is in the fault.

inSubfaults = sapply(1:nrow(faultGeom), getSubfaultGPSDat)

inFault = apply(inSubfaults, 1, any)

# return the GPS data within the fault

dat[inFault,]