1
|
################## Validation and analyses of results #######################################
|
2
|
############################ Covariate production for a given tile/region ##########################################
|
3
|
#This script examines inputs and outputs from the interpolation step.
|
4
|
#Part 1: Script produces plots for every selected date
|
5
|
#Part 2: Examine
|
6
|
#AUTHOR: Benoit Parmentier
|
7
|
#DATE: 02/22/2013
|
8
|
|
9
|
#PROJECT: NCEAS INPLANT: Environment and Organisms --TASK#???--
|
10
|
|
11
|
##Comments and TODO:
|
12
|
#Separate inteprolation results analyses from covariates analyses
|
13
|
|
14
|
##################################################################################################
|
15
|
|
16
|
###Loading R library and packages
|
17
|
library(RPostgreSQL)
|
18
|
library(sp) # Spatial pacakge with class definition by Bivand et al.
|
19
|
library(spdep) # Spatial pacakge with methods and spatial stat. by Bivand et al.
|
20
|
library(rgdal) # GDAL wrapper for R, spatial utilities
|
21
|
library(raster)
|
22
|
library(gtools)
|
23
|
library(rasterVis)
|
24
|
library(graphics)
|
25
|
library(grid)
|
26
|
library(lattice)
|
27
|
|
28
|
### Parameters and arguments
|
29
|
|
30
|
##Paths to inputs and output
|
31
|
in_path <- "/home/parmentier/Data/IPLANT_project/Venezuela_interpolation/Venezuela_01142013/input_data/"
|
32
|
out_path<- "/home/parmentier/Data/IPLANT_project/Venezuela_interpolation/Venezuela_01142013/output_data/"
|
33
|
infile3<-"covariates__venezuela_region__VE_01292013.tif" #this is an output from covariate script
|
34
|
|
35
|
setwd(in_path)
|
36
|
|
37
|
### Functions used in the script
|
38
|
|
39
|
load_obj <- function(f)
|
40
|
{
|
41
|
env <- new.env()
|
42
|
nm <- load(f, env)[1]
|
43
|
env[[nm]]
|
44
|
}
|
45
|
|
46
|
### PLOTTING RESULTS FROM VENEZUELA INTERPOLATION FOR ANALYSIS
|
47
|
|
48
|
#Select relevant dates and load R objects created during the interpolation step
|
49
|
|
50
|
date_selected<-c("20100103")
|
51
|
|
52
|
gam_fus_mod<-load_obj("gam_fus_mod_365d_GAM_fus5_all_lstd_02202013.RData")
|
53
|
validation_obj<-load_obj("gam_fus_validation_mod_365d_GAM_fus5_all_lstd_02202013.RData")
|
54
|
clim_obj<-load_obj("gamclim_fus_mod_365d_GAM_fus5_all_lstd_02202013.RData")
|
55
|
|
56
|
## Read covariate stack...
|
57
|
rnames <-c("x","y","lon","lat","N","E","N_w","E_w","elev","slope","aspect","CANHEIGHT","DISTOC")
|
58
|
lc_names<-c("LC1","LC2","LC3","LC4","LC5","LC6","LC7","LC8","LC9","LC10","LC11","LC12")
|
59
|
lst_names<-c("mm_01","mm_02","mm_03","mm_04","mm_05","mm_06","mm_07","mm_08","mm_09","mm_10","mm_11","mm_12",
|
60
|
"nobs_01","nobs_02","nobs_03","nobs_04","nobs_05","nobs_06","nobs_07","nobs_08",
|
61
|
"nobs_09","nobs_10","nobs_11","nobs_12")
|
62
|
|
63
|
covar_names<-c(rnames,lc_names,lst_names)
|
64
|
|
65
|
s_raster<-stack(infile3) #read in the data stack
|
66
|
names(s_raster)<-covar_names #Assigning names to the raster layers: making sure it is included in the extraction
|
67
|
|
68
|
## Figure 0: study area based on LC12 (water) mask
|
69
|
|
70
|
LC_mask<-subset(s_raster,"LC12")
|
71
|
LC_mask[LC_mask==100]<-NA
|
72
|
LC_mask <- LC_mask < 100
|
73
|
LC_mask_rec<-LC_mask
|
74
|
LC_mask_rec[is.na(LC_mask_rec)]<-0
|
75
|
|
76
|
png(paste("Study_area_",
|
77
|
out_prefix,".png", sep=""))
|
78
|
plot(LC_mask_rec,legend=FALSE,col=c("black","red"))
|
79
|
legend("topright",legend=c("Outside","Inside"),title="Study area",
|
80
|
pt.cex=0.9,fill=c("black","red"),bty="n")
|
81
|
title("Study area")
|
82
|
dev.off()
|
83
|
|
84
|
#determine index position matching date selected
|
85
|
|
86
|
i_dates<-vector("list",length(date_selected))
|
87
|
for (i in 1:length(gam_fus_mod)){
|
88
|
for (j in 1:length(date_selected)){
|
89
|
if(gam_fus_mod[[i]]$sampling_dat$date==date_selected[j]){
|
90
|
i_dates[[j]]<-i
|
91
|
}
|
92
|
}
|
93
|
}
|
94
|
|
95
|
#Examine the first select date add loop or function later
|
96
|
j=1
|
97
|
|
98
|
date<-strptime(date_selected[j], "%Y%m%d") # interpolation date being processed
|
99
|
month<-strftime(date, "%m") # current month of the date being processed
|
100
|
|
101
|
#Get raster stack of interpolated surfaces
|
102
|
index<-i_dates[[j]]
|
103
|
pred_temp<-as.character(gam_fus_mod[[index]]$dailyTmax) #list of files
|
104
|
rast_pred_temp<-stack(pred_temp) #stack of temperature predictions from models
|
105
|
|
106
|
#Get validation metrics, daily spdf training and testing stations, monthly spdf station input
|
107
|
sampling_dat<-gam_fus_mod[[index]]$sampling_dat
|
108
|
metrics_v<-validation_obj[[index]]$metrics_v
|
109
|
metrics_s<-validation_obj[[index]]$metrics_s
|
110
|
data_v<-validation_obj[[index]]$data_v
|
111
|
data_s<-validation_obj[[index]]$data_s
|
112
|
data_month<-clim_obj[[index]]$data_month
|
113
|
formulas<-clim_obj[[index]]$formulas
|
114
|
|
115
|
#Adding layer LST to the raster stack of covariates
|
116
|
#The names of covariates can be changed...
|
117
|
|
118
|
LST_month<-paste("mm_",month,sep="") # name of LST month to be matched
|
119
|
pos<-match("LST",layerNames(s_raster)) #Find the position of the layer with name "LST", if not present pos=NA
|
120
|
s_raster<-dropLayer(s_raster,pos) # If it exists drop layer
|
121
|
pos<-match(LST_month,layerNames(s_raster)) #Find column with the current month for instance mm12
|
122
|
r1<-raster(s_raster,layer=pos) #Select layer from stack
|
123
|
layerNames(r1)<-"LST"
|
124
|
#Get mask image!!
|
125
|
|
126
|
date_proc<-strptime(sampling_dat$date, "%Y%m%d") # interpolation date being processed
|
127
|
mo<-as.integer(strftime(date_proc, "%m")) # current month of the date being processed
|
128
|
day<-as.integer(strftime(date_proc, "%d"))
|
129
|
year<-as.integer(strftime(date_proc, "%Y"))
|
130
|
datelabel=format(ISOdate(year,mo,day),"%b %d, %Y")
|
131
|
|
132
|
## Figure 1: LST_TMax_scatterplot
|
133
|
|
134
|
rmse<-metrics_v$rmse[nrow(metrics_v)]
|
135
|
rmse_f<-metrics_s$rmse[nrow(metrics_s)]
|
136
|
|
137
|
png(paste("LST_TMax_scatterplot_",sampling_dat$date,"_",sampling_dat$prop,"_",sampling_dat$run_samp,
|
138
|
out_prefix,".png", sep=""))
|
139
|
plot(data_month$TMax,data_month$LST,xlab="Station mo Tmax",ylab="LST mo Tmax")
|
140
|
title(paste("LST vs TMax for",datelabel,sep=" "))
|
141
|
abline(0,1)
|
142
|
nb_point<-paste("n=",length(data_month$TMax),sep="")
|
143
|
mean_bias<-paste("Mean LST bias= ",format(mean(data_month$LSTD_bias,na.rm=TRUE),digits=3),sep="")
|
144
|
#Add the number of data points on the plot
|
145
|
legend("topleft",legend=c(mean_bias,nb_point),bty="n")
|
146
|
dev.off()
|
147
|
|
148
|
## Figure 2: Daily_tmax_monthly_TMax_scatterplot
|
149
|
|
150
|
png(paste("Daily_tmax_monthly_TMax_scatterplot_",sampling_dat$date,"_",sampling_dat$prop,"_",sampling_dat$run_samp,
|
151
|
out_prefix,".png", sep=""))
|
152
|
plot(dailyTmax~TMax,data=data_s,xlab="Mo Tmax",ylab=paste("Daily for",datelabel),main="across stations in VE")
|
153
|
nb_point<-paste("ns=",length(data_s$TMax),sep="")
|
154
|
nb_point2<-paste("ns_obs=",length(data_s$TMax)-sum(is.na(data_s[[y_var_name]])),sep="")
|
155
|
nb_point3<-paste("n_month=",length(data_month$TMax),sep="")
|
156
|
#Add the number of data points on the plot
|
157
|
legend("topleft",legend=c(nb_point,nb_point2,nb_point3),bty="n",cex=0.8)
|
158
|
dev.off()
|
159
|
|
160
|
## Figure 3: Predicted_tmax_versus_observed_scatterplot
|
161
|
|
162
|
#This is for mod_kr!! add other models later...
|
163
|
png(paste("Predicted_tmax_versus_observed_scatterplot_",sampling_dat$date,"_",sampling_dat$prop,"_",sampling_dat$run_samp,
|
164
|
out_prefix,".png", sep=""))
|
165
|
#plot(data_s$mod_kr~data_s[[y_var_name]],xlab=paste("Actual daily for",datelabel),ylab="Pred daily")
|
166
|
|
167
|
y_range<-range(c(data_s$mod_kr,data_v$mod_kr),na.rm=T)
|
168
|
x_range<-range(c(data_s[[y_var_name]],data_v[[y_var_name]]),na.rm=T)
|
169
|
col_t<- c("black","red")
|
170
|
pch_t<- c(1,2)
|
171
|
plot(data_s$mod_kr,data_s[[y_var_name]],
|
172
|
xlab=paste("Actual daily for",datelabel),ylab="Pred daily",
|
173
|
ylim=y_range,xlim=x_range,col=col_t[1],pch=pch_t[1])
|
174
|
points(data_v$mod_kr,data_v[[y_var_name]],col=col_t[2],pch=pch_t[2])
|
175
|
grid(lwd=0.5, col="black")
|
176
|
#plot(data_v$mod_kr~data_v[[y_var_name]],xlab=paste("Actual daily for",datelabel),ylab="Pred daily")
|
177
|
abline(0,1)
|
178
|
legend("topleft",legend=c("training","testing"),pch=pch_t,col=col_t,bty="n",cex=0.8)
|
179
|
title(paste("Predicted_tmax_versus_observed_scatterplot for",datelabel,sep=" "))
|
180
|
nb_point1<-paste("ns_obs=",length(data_s$TMax)-sum(is.na(data_s[[y_var_name]])),sep="")
|
181
|
rmse_str1<-paste("RMSE= ",format(rmse,digits=3),sep="")
|
182
|
rmse_str2<-paste("RMSE_f= ",format(rmse_f,digits=3),sep="")
|
183
|
|
184
|
#Add the number of data points on the plot
|
185
|
legend("bottomright",legend=c(nb_point1,rmse_str1,rmse_str2),bty="n",cex=0.8)
|
186
|
dev.off()
|
187
|
|
188
|
## Figure 5: prediction raster images
|
189
|
png(paste("Raster_prediction_",sampling_dat$date,"_",sampling_dat$prop,"_",sampling_dat$run_samp,
|
190
|
out_prefix,".png", sep=""))
|
191
|
#paste(metrics_v$pred_mod,format(metrics_v$rmse,digits=3),sep=":")
|
192
|
names(rast_pred_temp)<-paste(metrics_v$pred_mod,format(metrics_v$rmse,digits=3),sep=":")
|
193
|
#plot(rast_pred_temp)
|
194
|
levelplot(rast_pred_temp)
|
195
|
dev.off()
|
196
|
|
197
|
## Figure 6: training and testing stations used
|
198
|
png(paste("Training_testing_stations_map_",sampling_dat$date,"_",sampling_dat$prop,"_",sampling_dat$run_samp,
|
199
|
out_prefix,".png", sep=""))
|
200
|
plot(raster(rast_pred_temp,layer=5))
|
201
|
plot(data_s,col="black",cex=1.2,pch=2,add=TRUE)
|
202
|
plot(data_v,col="red",cex=1.2,pch=1,add=TRUE)
|
203
|
legend("topleft",legend=c("training stations", "testing stations"),
|
204
|
cex=1, col=c("black","red"),
|
205
|
pch=c(2,1),bty="n")
|
206
|
dev.off()
|
207
|
|
208
|
## Figure 7: monthly stations used
|
209
|
|
210
|
png(paste("Monthly_data_study_area",
|
211
|
out_prefix,".png", sep=""))
|
212
|
plot(raster(rast_pred_temp,layer=5))
|
213
|
plot(data_month,col="black",cex=1.2,pch=4,add=TRUE)
|
214
|
title("Monthly ghcn station in Venezuela for January")
|
215
|
dev.off()
|
216
|
|
217
|
## Figure 8: delta surface and bias
|
218
|
|
219
|
png(paste("Bias_delta_surface_",sampling_dat$date[i],"_",sampling_dat$prop[i],
|
220
|
"_",sampling_dat$run_samp[i],out_prefix,".png", sep=""))
|
221
|
|
222
|
bias_rast<-stack(clim_obj[[index]]$bias)
|
223
|
delta_rast<-raster(gam_fus_mod[[index]]$delta) #only one delta image!!!
|
224
|
names(delta_rast)<-"delta"
|
225
|
rast_temp_date<-stack(bias_rast,delta_rast)
|
226
|
rast_temp_date<-mask(rast_temp_date,LC_mask,file="test.tif",overwrite=TRUE)
|
227
|
#bias_d_rast<-raster("fusion_bias_LST_20100103_30_1_10d_GAM_fus5_all_lstd_02082013.rst")
|
228
|
plot(rast_temp_date)
|
229
|
|
230
|
dev.off()
|
231
|
|
232
|
#Figure 9: histogram for all images...
|
233
|
|
234
|
histogram(rast_pred_temp)
|
235
|
|
236
|
## Summarize information for the day: write out textfile...
|
237
|
|
238
|
#Number of station per month
|
239
|
#Number of station per day (training, testing,NA)
|
240
|
#metrics_v,metrics_s
|
241
|
#
|
242
|
|
243
|
# ################
|
244
|
# #PART 2: Region Covariate analyses ###
|
245
|
# ################
|
246
|
#
|
247
|
# # This should be in a separate script to analyze covariates from region.
|
248
|
#
|
249
|
# #MAP1:Study area with LC mask and tiles/polygon outline
|
250
|
#
|
251
|
#
|
252
|
# #MAP 2: plotting land cover in the study region:
|
253
|
#
|
254
|
# l1<-"LC1,Evergreen/deciduous needleleaf trees"
|
255
|
# l2<-"LC2,Evergreen broadleaf trees"
|
256
|
# l3<-"LC3,Deciduous broadleaf trees"
|
257
|
# l4<-"LC4,Mixed/other trees"
|
258
|
# l5<-"LC5,Shrubs"
|
259
|
# l6<-"LC6,Herbaceous vegetation"
|
260
|
# l7<-"LC7,Cultivated and managed vegetation"
|
261
|
# l8<-"LC8,Regularly flooded shrub/herbaceous vegetation"
|
262
|
# l9<-"LC9,Urban/built-up"
|
263
|
# l10<-"LC10,Snow/ice"
|
264
|
# l11<-"LC11,Barren lands/sparse vegetation"
|
265
|
# l12<-"LC12,Open water"
|
266
|
# lc_names_str<-c(l1,l2,l3,l4,l5,l6,l7,l8,l9,l10,l11,l12)
|
267
|
#
|
268
|
# names(lc_reg_s)<-lc_names_str
|
269
|
#
|
270
|
# png(paste("LST_TMax_scatterplot_",sampling_dat$date[i],"_",sampling_dat$prop[i],"_",sampling_dat$run_samp[i], out_prefix,".png", sep=""))
|
271
|
# plot(modst$TMax,sta_tmax_from_lst,xlab="Station mo Tmax",ylab="LST mo Tmax",main=paste("LST vs TMax for",datelabel,sep=" "))
|
272
|
# abline(0,1)
|
273
|
# nb_point<-paste("n=",length(modst$TMax),sep="")
|
274
|
# mean_bias<-paste("LST bigrasas= ",format(mean(modst$LSTD_bias,na.rm=TRUE),digits=3),sep="")
|
275
|
# #Add the number of data points on the plot
|
276
|
# legend("topleft",legend=c(mean_bias,nb_point),bty="n")
|
277
|
# dev.off()
|
278
|
#
|
279
|
# #Map 3: Elevation and LST in January
|
280
|
# tmp_s<-stack(LST,elev_1)
|
281
|
# png(paste("LST_elev_",sampling_dat$date[i],"_",sampling_dat$prop[i],"_",sampling_dat$run_samp[i], out_prefix,".png", sep=""))
|
282
|
# plot(tmp_s)
|
283
|
#
|
284
|
# #Map 4: LST climatology per month
|
285
|
#
|
286
|
# names_tmp<-c("mm_01","mm_02","mm_03","mm_04","mm_05","mm_06","mm_07","mm_08","mm_09","mm_10","mm_11","mm_12")
|
287
|
# LST_s<-subset(s_raster,names_tmp)
|
288
|
# names_tmp<-c("nobs_01","nobs_02","nobs_03","nobs_04","nobs_05","nobs_06","nobs_07","nobs_08",
|
289
|
# "nobs_09","nobs_10","nobs_11","nobs_12")
|
290
|
# LST_nobs<-subset(s_raster,names_tmp)
|
291
|
#
|
292
|
# LST_nobs<-mask(LST_nobs,LC_mask,filename="test2.tif")
|
293
|
# LST_s<-mask(LST_s,LC_mask,filename="test3.tif")
|
294
|
# c("Jan","Feb")
|
295
|
# plot(LST_s)
|
296
|
# plot(LST_nobs)
|
297
|
#
|
298
|
# #Map 5: LST and TMax
|
299
|
#
|
300
|
# #note differnces in patternin agricultural areas and
|
301
|
# min_values<-cellStats(LST_s,"min")
|
302
|
# max_values<-cellStats(LST_s,"max")
|
303
|
# mean_values<-cellStats(LST_s,"mean")
|
304
|
# sd_values<-cellStats(LST_s,"sd")
|
305
|
# #median_values<-cellStats(molst,"median") Does not extist
|
306
|
# statistics_LST_s<-cbind(min_values,max_values,mean_values,sd_values) #This shows that some values are extremes...especially in October
|
307
|
# LST_stat_data<-as.data.frame(statistics_LST_s)
|
308
|
# names(LST_stat_data)<-c("min","max","mean","sd")
|
309
|
# # Statistics for number of valid observation stack
|
310
|
# min_values<-cellStats(nobslst,"min")
|
311
|
# max_values<-cellStats(nobslst,"max")
|
312
|
# mean_values<-cellStats(nobslst,"mean")
|
313
|
# sd_values<-cellStats(nobslst,"sd")
|
314
|
# statistics_LSTnobs_s<-cbind(min_values,max_values,mean_values,sd_values) #This shows that some values are extremes...especially in October
|
315
|
# LSTnobs_stat_data<-as.data.frame(statistics_LSTnobs_s)
|
316
|
#
|
317
|
# X11(width=12,height=12)
|
318
|
# #Plot statiscs (mean,min,max) for monthly LST images
|
319
|
# plot(1:12,LST_stat_data$mean,type="b",ylim=c(-15,60),col="black",xlab="month",ylab="tmax (degree C)")
|
320
|
# lines(1:12,LST_stat_data$min,type="b",col="blue")
|
321
|
# lines(1:12,LST_stat_data$max,type="b",col="red")
|
322
|
# text(1:12,LST_stat_data$mean,rownames(LST_stat_data),cex=1,pos=2)
|
323
|
#
|
324
|
# legend("topleft",legend=c("min","mean","max"), cex=1.5, col=c("blue","black","red"),
|
325
|
# lty=1)
|
326
|
# title(paste("LST statistics for Oregon", "2010",sep=" "))
|
327
|
# savePlot("lst_statistics_OR.png",type="png")
|
328
|
#
|
329
|
# #Plot number of valid observations for LST
|
330
|
# plot(1:12,LSTnobs_stat_data$mean,type="b",ylim=c(0,280),col="black",xlab="month",ylab="tmax (degree C)")
|
331
|
# lines(1:12,LSTnobs_stat_data$min,type="b",col="blue")
|
332
|
# lines(1:12,LSTnobs_stat_data$max,type="b",col="red")
|
333
|
# text(1:12,LSTnobs_stat_data$mean,rownames(LSTnobs_stat_data),cex=1,pos=2)
|
334
|
#
|
335
|
# legend("topleft",legend=c("min","mean","max"), cex=1.5, col=c("blue","black","red"),
|
336
|
# lty=1)
|
337
|
# title(paste("LST number of valid observations for Oregon", "2010",sep=" "))
|
338
|
# savePlot("lst_nobs_OR.png",type="png")
|
339
|
#
|
340
|
# plot(data_month$TMax,add=TRUE)
|
341
|
#
|
342
|
# ### Map 6: station in the region
|
343
|
#
|
344
|
# plot(tmax_predicted)
|
345
|
# plot(data_s,col="black",cex=1.2,pch=4,add=TRUE)
|
346
|
# plot(data_v,col="blue",cex=1.2,pch=2,add=TRUE)
|
347
|
#
|
348
|
# plot(tmax_predicted)
|
349
|
# plot(data_month,col="black",cex=1.2,pch=4,add=TRUE)
|
350
|
# title("Monthly ghcn station in Venezuela for 2000-2010")
|
351
|
#
|
352
|
#png...output?
|
353
|
# plot(interp_area, axes =TRUE)
|
354
|
# plot(stat_reg, pch=1, col="red", cex= 0.7, add=TRUE)
|
355
|
# plot(data_reg,pch=2,col="blue",cex=2,add=TRUE)
|
356
|
|
357
|
#### End of script ####
|