forked from rodaguayo/Glacier_Uncertainties
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathPreprocessing_catchments.R
228 lines (179 loc) · 12 KB
/
Preprocessing_catchments.R
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
# Code for RGI and catchment preprocessing ------------------------------------------------------
# Developed by Rodrigo Aguayo (2022-2023)
rm(list=ls())
cat("\014")
setwd("/home/rooda/Dropbox/Patagonia/")
#setwd("C:/Users/rooda/Dropbox/Patagonia/")
library("exactextractr")
library("rgrass")
library("terra")
library("stats")
library("dplyr")
library("sf")
sf_use_s2(FALSE)
# 1. Delimitation of all basins -------------------------------------------------------------------
dem <- rast("GIS South/dem_patagonia3f.tif") # DEM from NASADEM 90m
dem <- aggregate(dem, fact=2, fun="mean")
dem <- crop(dem, ext(c(-75, -67, -56, -40)))
dem <- project(dem, "EPSG:32719")
initGRASS(gisBase = "/usr/lib/grass82/", home="/home/rooda/Dropbox/Patagonia/",
SG = dem, override = TRUE)
write_RAST(dem, vname = "dem_grass", flags = c("overwrite", "o"))
write_RAST(is.na(dem), vname = "depre_grass", flags = c("overwrite", "o")) # depression DEM -> water/ocean
execGRASS("r.stream.extract", flags=c("overwrite"),
parameters = list(elevation="dem_grass", depression = "depre_grass", threshold = 200,
direction = "fdir", stream_vector="stream_v", stream_raster="stream_r"))
execGRASS("r.stream.basins", flags=c("overwrite", "l"),
parameters=list(direction="fdir", stream_rast = "stream_r", basins="basins"))
# 1.1 Attributes and subset by area and location --------------------------------------------------
all_basins <- read_RAST("basins")
all_basins <- as.polygons(all_basins)
all_basins <- project(all_basins, "EPSG:4326")
all_basins$basin_area <- round(expanse(all_basins, unit="km"), 2)
all_basins$lat <- as.data.frame(geom(centroids(all_basins)))$y
all_basins$lon <- as.data.frame(geom(centroids(all_basins)))$x
all_basins <- subset(all_basins, (all_basins$lat < -41) & (all_basins$lat > -55.5))
all_basins <- subset(all_basins, (all_basins$lon < -70) | (all_basins$lat < -54))
all_basins <- subset(all_basins, !((all_basins$lon > -71) & (all_basins$lat > -50) & (all_basins$lat < -48)))
all_basins <- subset(all_basins, (all_basins$lon < -68.5))
all_basins <- subset(all_basins, all_basins$basin_area > 10)
# 1.2 Assign Zone and preliminary ID to all basins ------------------------------------------------
all_basins$Zone <- 0 # Initialization
all_basins$Zone <- ifelse(all_basins$lat > -43.4, 1, all_basins$Zone)
all_basins$Zone <- ifelse(all_basins$lat < -43.4 & all_basins$lat > -46, 2, all_basins$Zone)
all_basins$Zone <- ifelse(all_basins$lat < -46 & all_basins$lat > -47.8 & all_basins$lon > -73, 3, all_basins$Zone)
all_basins$Zone <- ifelse(all_basins$lat < -46 & all_basins$lat > -47.8 & all_basins$lon < -73, 4, all_basins$Zone)
all_basins$Zone <- ifelse(all_basins$lat < -47.8 & all_basins$lat > -49.4, 5, all_basins$Zone)
all_basins$Zone <- ifelse(all_basins$lat < -49.4 & all_basins$lat > -50.7, 6, all_basins$Zone)
all_basins$Zone <- ifelse(all_basins$lat < -50.7 & all_basins$lat > -52.1 , 7, all_basins$Zone)
all_basins$Zone <- ifelse(all_basins$lat < -52.1 & all_basins$lat > -54.1 , 8, all_basins$Zone)
all_basins$Zone <- ifelse(all_basins$lat < -54.1, 9, all_basins$Zone)
all_basins$ID <- seq(0, nrow(all_basins)-1) # preliminary IDs
plot(all_basins, "Zone")
all_basins <- st_as_sf(all_basins)
all_basins <- st_transform(all_basins, 32718) # UTM 18S for the glaciers
# delete trash from GRASS (file12135term1k2 etc)
all_basins <- all_basins[,c("basin_area", "lat", "lon", "ID", "Zone", "geometry")]
# 2. RGI glaciers: Selection and zone assignment --------------------------------------------------
RGI <- list(st_read("GIS South/Glaciers/RGI6.shp"), st_read("GIS South/Glaciers/RGI7.shp"))
dem <- rast("GIS South/dem_patagonia3f.tif")
dem <- project(dem, "epsg:32718", method = "bilinear")
dem <- subst(dem, NA, 0) # NAs to sea level (= 0)
for (i in 1:2) {
RGI_i <- subset(RGI[[i]], RGI[[i]]$CenLat < -40.5)
RGI_i <- st_transform(RGI_i, 32718) # UTM 18S
# terminus location
RGI_ic <- extract(dem, as.lines(vect(RGI_i)), xy = TRUE)
RGI_ic <- RGI_ic %>% group_by(ID) %>% slice_min(order_by = dem_patagonia3f)
RGI_ic <- aggregate(RGI_ic, by = list(RGI_ic$ID), FUN = mean) # several options if terminus close to water
RGI_ic <- st_as_sf(RGI_ic, coords = c("x","y"), crs = 32718)
# intersection with basins
RGI_ic <- st_join(RGI_ic["ID"], all_basins[,c("Zone", "ID", "geometry")], join = st_within)
RGI_i$Zone <- RGI_ic$Zone # Assign zone based on terminus location
RGI_i$ID_basin <- RGI_ic$ID.y # Assign catchment
# centroid location if terminus is outside the basins
RGI_ic <- st_centroid(RGI_i["geometry"])
RGI_ic <- st_join(RGI_ic, all_basins[,c("Zone", "ID", "geometry")], join = st_within)
RGI_i$Zone[is.na(RGI_i$Zone)] <- RGI_ic$Zone[is.na(RGI_i$Zone)]
RGI_i$ID_basin[is.na(RGI_i$ID_basin)] <- RGI_ic$ID[is.na(RGI_i$ID_basin)] # Assign catchment
# select glaciers with a code for the Zone
RGI_i <- subset(RGI_i, RGI_i$Zone > 0)
RGI_i$O2Region <- 1 # useful to count
RGI_i$area_km2 <- expanse(vect(RGI_i), unit="km")
RGI[[i]] <- RGI_i
}
# 3. Subset catchments if there are enough glaciers -----------------------------------------------
# if there at least one glacier in RGI6 or RGI7
RGI_groupby <- list(aggregate(st_drop_geometry(RGI[[1]][,c("O2Region", "area_km2")]),
by = list(RGI[[1]]$ID_basin), FUN = "sum"),
aggregate(st_drop_geometry(RGI[[2]][,c("O2Region", "area_km2")]),
by = list(RGI[[2]]$ID_basin), FUN = "sum"))
RGI_groupby <- merge(RGI_groupby[[1]], RGI_groupby[[2]], by = "Group.1", all=TRUE)
RGI_groupby[is.na(RGI_groupby)] <- 0
colnames(RGI_groupby) <- c("ID", "RGI6_ngla","RGI6_area","RGI7_ngla", "RGI7_area")
all_basins <- subset(all_basins, all_basins$ID %in% RGI_groupby$ID)
# if the glacier area in RGI6 or RGI7 is higher than 0.1%
all_basins <- merge(all_basins, RGI_groupby, by = "ID", all=TRUE)
all_basins <- subset(all_basins, all_basins$RGI6_area/all_basins$basin_area > 0.001 |
all_basins$RGI7_area/all_basins$basin_area > 0.001)
# 4. Subset glaciers again ------------------------------------------------------------------------
RGI[[1]] <- subset(RGI[[1]], RGI[[1]]$ID_basin %in% all_basins$ID) # small effect
RGI[[2]] <- subset(RGI[[2]], RGI[[2]]$ID_basin %in% all_basins$ID)
# 5. Finish catchment file ------------------------------------------------------------------------
glacier_area <- vect("GIS South/Glaciers/RGI6.shp") # gridded area to compare just in case
glacier_area <- rasterize(glacier_area, project(dem, "EPSG:4326"), background = 0) * 100
all_basins <- st_transform(all_basins, 4326) # go back to WGS84
all_basins$RGI6_area_r <- extract(glacier_area, all_basins, "mean")[,2]
writeVector(vect(all_basins), "GIS South/Basins_Patagonia_all.shp", overwrite=TRUE)
plot(all_basins) # visual check
# 6. Glacier attributes ---------------------------------------------------------------------------
## 6.1 Topography ----------------------------------------------------------------------------------
slope <- terrain(dem, v="slope", neighbors=8, unit="degrees")
aspect <- terrain(dem, v="aspect", neighbors=8, unit="degrees")
for (i in 1:2) {
RGI[[i]]$slope_v2 <- exact_extract(slope, RGI[[i]], "mean")
RGI[[i]]$aspect_v2 <- exact_extract(aspect, RGI[[i]], "mean")
}
## 6.2 Volume data from Farinotti et al. and Millan et al. ----------------------------------------
vol_F19 <- rast("GIS South/Glaciers/Thickness_2019.tif")
vol_M22 <- rast("GIS South/Glaciers/Thickness_2022.tif")
### 6.2.1 Millan et al. 2022 (M22) ---------------------------------------------------------------
RGI6 <- RGI[[1]]
RGI7 <- RGI[[2]]
RGI6$vol_M22 <- exact_extract(vol_M22, RGI6, "sum") * res(vol_M22)[1] * res(vol_M22)[2] * 1e-9
RGI7$vol_M22 <- exact_extract(vol_M22, RGI7, "sum") * res(vol_M22)[1] * res(vol_M22)[2] * 1e-9
RGI6$vol_M22c <- round(exact_extract(not.na(vol_M22), RGI6, "mean"), 3)
RGI7$vol_M22c <- round(exact_extract(not.na(vol_M22), RGI7, "mean"), 3)
RGI6$vol_M22[RGI6$vol_M22c < 0.5] <- NA
RGI7$vol_M22[RGI7$vol_M22c < 0.5] <- NA
for (i in sort(unique(RGI6$Zone))) { # VAS based on M22 data
model6 <- lm(log(RGI6$vol_M22)[RGI6$Zone == i & RGI6$vol_M22c > 0.9] ~ log(RGI6$area_km2)[RGI6$Zone == i & RGI6$vol_M22c > 0.9] )
model7 <- lm(log(RGI7$vol_M22)[RGI7$Zone == i & RGI7$vol_M22c > 0.9] ~ log(RGI7$area_km2)[RGI7$Zone == i & RGI7$vol_M22c > 0.9] )
RGI6$vol_M22[RGI6$Zone == i & is.na(RGI6$vol_M22)] <- exp(coef(model6)[1]) * RGI6$area_km2[RGI6$Zone == i & is.na(RGI6$vol_M22)] ** coef(model6)[2]
RGI7$vol_M22[RGI7$Zone == i & is.na(RGI7$vol_M22)] <- exp(coef(model7)[1]) * RGI7$area_km2[RGI7$Zone == i & is.na(RGI7$vol_M22)] ** coef(model7)[2]
}
### 6.2.2 Farinotti et al. 2019 (F19) ------------------------------------------------------------
RGI6$vol_F19 <- exact_extract(vol_F19, RGI6, "sum") * res(vol_F19)[1] * res(vol_F19)[2] * 1e-9
RGI7$vol_F19 <- exact_extract(vol_F19, RGI7, "sum") * res(vol_F19)[1] * res(vol_F19)[2] * 1e-9
RGI6$vol_F19c <- round(exact_extract(not.na(vol_F19), RGI6, "mean"), 3)
RGI7$vol_F19c <- round(exact_extract(not.na(vol_F19), RGI7, "mean"), 3)
RGI6$vol_F19[RGI6$vol_F19c < 0.5] <- NA
RGI7$vol_F19[RGI7$vol_F19c < 0.5] <- NA
RGI6$vol_F19[RGI6$vol_F19 == 0] <- 1e-6 # few glaciers with 0 volume (problem for log scale)
RGI7$vol_F19[RGI7$vol_F19 == 0] <- 1e-6
for (i in sort(unique(RGI6$Zone))) { # VAS based on F19 data
model6 <- lm(log(RGI6$vol_F19)[RGI6$Zone == i & RGI6$vol_F19c > 0.9] ~ log(RGI6$area_km2)[RGI6$Zone == i & RGI6$vol_F19c > 0.9] )
model7 <- lm(log(RGI7$vol_F19)[RGI7$Zone == i & RGI7$vol_F19c > 0.9] ~ log(RGI7$area_km2)[RGI7$Zone == i & RGI7$vol_F19c > 0.9] )
RGI6$vol_F19[RGI6$Zone == i & is.na(RGI6$vol_F19)] <- exp(coef(model6)[1]) * RGI6$area_km2[RGI6$Zone == i & is.na(RGI6$vol_F19)] ** coef(model6)[2]
RGI7$vol_F19[RGI7$Zone == i & is.na(RGI7$vol_F19)] <- exp(coef(model7)[1]) * RGI7$area_km2[RGI7$Zone == i & is.na(RGI7$vol_F19)] ** coef(model7)[2]
}
## 6.3 dmdtda: specific-mass change rate in meters water-equivalent per year ----------------------
dhdt_21 <- rast("GIS South/Glaciers/dhdt_2021.tif")
dhdt_source <- read.csv("MS2 Results/dhdt_origin_theia.csv")
dhdt_source <- subset(dhdt_source, dhdt_source$period == "2000-01-01_2020-01-01")
dhdt_source <- dhdt_source[dhdt_source$rgiid %in% RGI6$RGIId,] # same order
RGI6$dmdtda_21 <- exact_extract(dhdt_21, RGI6, "mean") * 0.850 # dhdt to dmdtda
RGI7$dmdtda_21 <- exact_extract(dhdt_21, RGI7, "mean") * 0.850
RGI6$dmdtda_21c <- round(exact_extract(not.na(dhdt_21), RGI6, "mean"), 3)
RGI7$dmdtda_21c <- round(exact_extract(not.na(dhdt_21), RGI7, "mean"), 3)
RGI6$dmdtda_21[RGI6$dmdtda_21c < 0.9] <- NA
RGI7$dmdtda_21[RGI7$dmdtda_21c < 0.9] <- NA
RGI6$dmdtda_error <- dhdt_source$err_dmdtda
## 6.3.1 Filling: Every glacier needs to have a dhdt (dmdadt) ------------------------------------
dmdtda_RGI6_mean <- sapply(split(RGI6, RGI6$Zone), function(d) weighted.mean(d$dmdtda_21, w = d$area_km2, na.rm = T)) # area-weighted is better :)
dmdtda_RGI7_mean <- sapply(split(RGI7, RGI7$Zone), function(d) weighted.mean(d$dmdtda_21, w = d$area_km2, na.rm = T))
for (i in sort(unique(RGI6$Zone))) {
RGI6$dmdtda_21[RGI6$Zone == i & is.na(RGI6$dmdtda_21)] <- dmdtda_RGI6_mean[i]
RGI7$dmdtda_21[RGI7$Zone == i & is.na(RGI7$dmdtda_21)] <- dmdtda_RGI7_mean[i]
}
# 7. Save glacier file ----------------------------------------------------------------------------
# go back to WGS84 (RGI format)
RGI6 <- st_transform(RGI6, 4326)
RGI7 <- st_transform(RGI7, 4326)
# area is area_km2
RGI7$area <- NULL
# source year in RGI format
RGI7$src_date <- paste0(substr(RGI7$src_date, 0,4), substr(RGI7$src_date, 6,7), substr(RGI7$src_date, 9,10))
# save using terra (problems in sf)
writeVector(vect(RGI6), "GIS South/Glaciers/RGI6_v2.shp", overwrite=TRUE)
writeVector(vect(RGI7), "GIS South/Glaciers/RGI7_v2.shp", overwrite=TRUE)