-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathprocess_HERMES_Chl_v2.R
1130 lines (971 loc) · 49.2 KB
/
process_HERMES_Chl_v2.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
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
#
# url<- "ftp://ftp.hermes.acri.fr/345576024"
# filenames <- getURL(url, userpwd="ftp_hermes:hermes%", ftp.use.epsv = FALSE, dirlistonly = TRUE) #reading filenames from ftp-server
# OCCI https://esa-oceancolour-cci.org/
# https://rsg.pml.ac.uk/thredds/ncss/grid/CCI_ALL-v4.0-8DAY/dataset.html
# NCSS Request URL
# https://rsg.pml.ac.uk/thredds/ncss/grid/CCI_ALL-v4.0-8DAY/dataset.html
# /thredds/ncss/CCI_ALL-v4.0-8DAY
# https://rsg.pml.ac.uk/thredds/ncss/CCI_ALL-v4.0-8DAY?var=chlor_a&north=48&west=-80&east=-60&south=32&horizStride=1&time_start=1997-09-04T00%3A00%3A00Z&time_end=2018-12-27T00%3A00%3A00Z&timeStride=1&addLatLon=true
#get OCCI v4.2
gribfile='http://rsg.pml.ac.uk/thredds/ncss/CCI_ALL-v4.2-8DAY?var=chlor_a&var=chlor_a_log10_bias&var=chlor_a_log10_rmsd&north=48&west=-76&east=-64&south=35&disableProjSubset=on&horizStride=1&time_start=1997-09-04T00%3A00%3A00Z&time_end=2019-12-27T00%3A00%3A00Z&timeStride=1&addLatLon=true'
download.file(gribfile,'junk.nc',mode = "wb")
library(raster)
library(mgcv)
# library(sp)
library(maptools)
library(marmap)
library(maps)
library(mapdata)
library(rgeos)
library(ncdf4)
library(abind)
library(RColorBrewer)
library(chron)
### Function to load ncdf files as stacked raster
nc2raster=function(x, varn){
s=stack()
yy=length(x)
for (i in 1:length(x)){
r <- raster(x[i], varname = varn)
s=stack(s, r)
print(paste(i,' of ', yy, ' files', sep=''))
}
return(s)
}
### Function to extract data using a shapefile
extract_calc=function(x, shp){
v2=list()
for(i in 1:dim(x)[3]){
v=extract(x[[i]], shp)
v1=lapply(v, function(xx) mean(xx, na.rm=T))
v2[i]=list(v1)
}
m=matrix(unlist(v2), ncol=dim(x)[3], nrow=length(shp@polygons)) # box 0-29 =rows, years =cols
# colnames(m)=seq(1998, 2016, by=1)
# rownamse(m)=
return(m)
}
### Function to plot a raster for the NES area and set scale
plotChlRaster=function(data, i, maxV, limit=F, dateval){
rasterX=data[[i]]
rng2=cellStats(rasterX, range)
if (limit == 1){
max_abolute_value=maxV #set limit manually with maxV input
rng=c(0, max_abolute_value, rng2[2])
}
else {
max_abolute_value=ceiling(rng2[2]) #round up actual max val
rng=c(0, max_abolute_value)
}
color=rev(brewer.pal(11, "Spectral"))
br <- seq(0, max_abolute_value, length.out=9)
arg=list(at=rng, labels=round(rng,1))
plot(rasterX, col=color, breaks=br,axis.args=arg, xlim=c(-77,-64),ylim=c(35,45),
las=1, legend=F, main=dateval[[i]])
map("worldHires", xlim=c(-77,-64),ylim=c(35,45), fill=T,border=0,col="gray", add=T)
plot(rasterX, legend.only=T, col=color,breaks=br,axis.args=arg, legend.shrink=0.5,
smallplot=c(0.19,0.21, 0.6,0.80) )
}
# color.bar <- function(lut, min, max=-min, nticks=11, ticks=seq(min, max, len=nticks), title='') {
# scale = (length(lut)-1)/(max-min)
#
# dev.new(width=1.75, height=5)
# plot(c(0,10), c(min,max), type='n', bty='n', xaxt='n', xlab='', yaxt='n', ylab='', main=title)
# axis(2, ticks, las=1)
# for (i in 1:(length(lut)-1)) {
# y = (i-1)/scale + min
# rect(0,y,10,y+1/scale, col=lut[i], border=NA)
# }
# }
#
# plot_binned_strata=function(m, i, shp, maxV, limit=F){
# rng2=cellStats(rasterX, range)
# if (limit == 1){
# max_abolute_value=maxV #set limit manually with maxV input
# rng=c(0, max_abolute_value, rng2[2])
# }
# else {
# max_abolute_value=ceiling(rng2[2]) #round up actual max val
# rng=c(0, max_abolute_value)
# }
# color=rev(brewer.pal(11, "Spectral"))
# br <- seq(0, max_abolute_value, length.out=length(color))
# m.bin=cut(m, br)
# arg=list(at=rng, labels=round(rng,1))
# plot(shp, col=color[m.bin], breaks=br,axis.args=arg, xlim=c(-77,-64),ylim=c(35,45),
# las=1, legend=F, main=av.dates[[i]])
# map("worldHires", xlim=c(-77,-64),ylim=c(35,45), fill=T,border=0,col="gray", add=T)
# color.bar(color, 0,max_abolute_value)
# # plot(rasterX, legend.only=T, col=color,breaks=br,axis.args=arg, legend.shrink=0.5,
# # smallplot=c(0.19,0.21, 0.6,0.80) )
# }
# # testing...
# plotChlRaster(chl.av, 5, 5, limit=T)
# plotChlRaster(chl.av, 5, 20, limit=F)
# ## load shapefiles
# setwd("G:/1 RM/KINGSTON/transfer/shapefiles/epu_shapes")
setwd("C:/Users/ryan.morse/Desktop/Iomega Drive Backup 20171012/1 RM/KINGSTON/transfer/shapefiles/epu_shapes")
d2='/home/ryan/Desktop/shapefiles/epu_shapes'
wd=getwd()
gbk=rgdal::readOGR(paste(d2, '/', "EPU_GBKPoly.shp", sep=''))
gom=rgdal::readOGR(paste(d2, '/', "EPU_GOMPoly.shp", sep=''))
mab=rgdal::readOGR(paste(d2, '/', "EPU_MABPoly.shp", sep=''))
scs=rgdal::readOGR(paste(d2, '/', "EPU_SCSPoly.shp", sep=''))
#combine shapefiles GOM and GBK
gom.gbk.shp=gUnion(gom, gbk, byid=F, id=NULL)
gom.gbk.shp=gUnion(gom, gbk, byid=F, id=NULL)
gom.scs.shp=gUnion(gom, scs, byid=F, id=NULL)
mab.gbk.shp=gUnion(mab, gbk, byid=F, id=NULL)
NES.shp=gUnion(mab.gbk.shp, gom.scs.shp, byid=F, id=NULL)
setwd("C:/Users/ryan.morse/Desktop/NES_5area")
nes.five=rgdal::readOGR('nes_gbk_gome_gomw_mabn_mabsPoly.shp')
## grab ecomon strata and plot
setwd('C:/Users/ryan.morse/Desktop/Iomega Drive Backup 20171012/1 RM/2 Plankton Spatial Plots/shapefiles')
ecomon.strata=rgdal::readOGR("EcoMon_strata.shp")
par(mar = c(0,0,0,0))
par(oma = c(0,0,0,0))
pdf(file='EcoMon_strata.pdf')
map("worldHires", xlim=c(-77,-65),ylim=c(35,45), fill=T,border=0,col="gray70")
map.axes(las=1)
data(stateMapEnv)
map('state', fill = F, add=T) # add state lines
lines(ecomon.strata)
text(coordinates(ecomon.strata)[,1], coordinates(ecomon.strata)[,2], ecomon.strata$STRATA)
dev.off()
#### Load Chl data
# AVW: weighted average of single-sensor Level 2 CHL1 products
# GSM: GSM merging of single sensor L3 NRRS
# The CHL1 algorithms are applicable for "case 1" waters
# OC5 Gohin, F., 2011
# CHL2 is the chlorophyll concentration (mg/m3) for Case 2 waters (see section validity); L3 merge: AV; sensors: MER, OLA; Doerffer and Schiller (2007)
# CHL2 uses the a Neural Network algorithm;The product is valid for case 2 waters, i.e. waters where inorganic particles dominate over phytoplankton (typically in coastal waters).
### Get 8-day OCCI chlorophyll data
setwd('/media/ryan/Iomega_HDD/1 RM/3 gridded data/OCCI')
setwd('C:/Users/ryan.morse/Desktop/Iomega Drive Backup 20171012/1 RM/3 gridded data/OCCI')
setwd('H:/1 RM/3 gridded data/OCCI')
# nc1=nc_open('CCI_ALL-v4.0-8DAY.nc') # just chl
# nc1=nc_open('C:/Users/ryan.morse/Downloads/CCI_ALL-v4.0-8DAY.nc') # new file with error estimates
nc1=nc_open('C:/Users/ryan.morse/Documents/GitHub/JPSS/CCI_ALL-v4.2-8DAY.nc') #udpated with data fix 2019
lon.occi=ncvar_get(nc1, 'lon')
lat.occi=ncvar_get(nc1, 'lat')
chl.occi=ncvar_get(nc1, 'chlor_a')
chl.occi.bias=ncvar_get(nc1, 'chlor_a_log10_bias') # only if using new file in downloads (2GB)
chl.occi.rmsd=ncvar_get(nc1, 'chlor_a_log10_rmsd') # only if using new file in downloads (2GB)
time.occi=ncvar_get(nc1, 'time') #days since Jan 1, 1970
dim(chl.occi)
colnames(chl.occi)=lat.occi
rownames(chl.occi)=lon.occi
nc_close(nc1)
test=month.day.year(time.occi, c(1,1,1970)) # these are not 8-days apart.... something odd
occi.date=data.frame(test)
occi.date$F1=paste(occi.date[,3], occi.date[,1], occi.date[,2], sep='-')
occi.date$DOY=as.numeric(strftime(occi.date$F1, '%j'))
ddiff=diff(occi.date$DOY)
occi.date$diff=c(0, ddiff) # see OC-CCI manual in JPSS/calibration folder for list of missing dates, explains why some are not 8d
table(occi.date$year)
occi.date$week=c(seq(from=31, to=46, by=1), rep(seq(from=1, to=46, by=1), 21))
# m=(t(chl.occi[,,3]))
# dimnames(m) <- list(lat=as.numeric(lat.occi), lon=as.numeric(lon.occi))
# t=raster(m)
# extent(t)=c(-80, -60, 32, 48)
# crs(t)="+proj=longlat +datum=WGS84 +no_defs +ellps=WGS84 +towgs84=0,0,0"
# plot(t)
# occi=brick((chl.occi))
# extent(t)=c(-80, -60, 32, 48)
# crs(t)="+proj=longlat +datum=WGS84 +no_defs +ellps=WGS84 +towgs84=0,0,0"
## Chl_loop over and stack
bb=c(-80, -60, 32, 48)
m2=t(chl.occi[,,1])
# m2=t(m)#[ncol(m):1,] # flip and transpose matrix
occi=raster(m2)
extent(occi)=bb
crs(occi)="+proj=longlat +datum=WGS84 +no_defs +ellps=WGS84 +towgs84=0,0,0"
for(i in 2:dim(chl.occi)[3]){
m2=t(chl.occi[,,i])
# m2=t(m)#[ncol(m):1,] # flip and transpose matrix
xx=raster(m2)
extent(xx)=bb
crs(xx)="+proj=longlat +datum=WGS84 +no_defs +ellps=WGS84 +towgs84=0,0,0"
# plot(xx)
occi=stack(occi, xx)
print(i)
}
## Chl_RMSD loop over and stack
bb=c(-80, -60, 32, 48)
m2=t(chl.occi.rmsd[,,1])
# m2=t(m)#[ncol(m):1,] # flip and transpose matrix
occi.rmsd=raster(m2)
extent(occi.rmsd)=bb
crs(occi.rmsd)="+proj=longlat +datum=WGS84 +no_defs +ellps=WGS84 +towgs84=0,0,0"
for(i in 2:dim(chl.occi.rmsd)[3]){
m2=t(chl.occi.rmsd[,,i])
# m2=t(m)#[ncol(m):1,] # flip and transpose matrix
xx=raster(m2)
extent(xx)=bb
crs(xx)="+proj=longlat +datum=WGS84 +no_defs +ellps=WGS84 +towgs84=0,0,0"
# plot(xx)
occi.rmsd=stack(occi.rmsd, xx)
print(i)
}
## Chl_Bias loop over and stack
bb=c(-80, -60, 32, 48)
m2=t(chl.occi.bias[,,1])
# m2=t(m)#[ncol(m):1,] # flip and transpose matrix
occi.bias=raster(m2)
extent(occi.bias)=bb
crs(occi.bias)="+proj=longlat +datum=WGS84 +no_defs +ellps=WGS84 +towgs84=0,0,0"
for(i in 2:dim(chl.occi.bias)[3]){
m2=t(chl.occi.bias[,,i])
# m2=t(m)#[ncol(m):1,] # flip and transpose matrix
xx=raster(m2)
extent(xx)=bb
crs(xx)="+proj=longlat +datum=WGS84 +no_defs +ellps=WGS84 +towgs84=0,0,0"
# plot(xx)
occi.bias=stack(occi.bias, xx)
print(i)
}
### HERMES 8-day composites ###
# setwd('G:/1 RM/3 gridded data/HERMES merged CHL 25km')
setwd("C:/Users/ryan.morse/Desktop/Iomega Drive Backup 20171012/1 RM/3 gridded data/HERMES_8day")
setwd('/media/ryan/Iomega_HDD/1 RM/3 gridded data/HERMES_8day')
wd=getwd()
## HERMES chl1 product for class 1 ocean waters, inlcudes GSM, AV, and AVW files
files.chl1=list.files(wd, pattern=('_CHL1_8D'))
files.chl1.ave=files.chl1[grep(files.chl1, pattern=('_AV-'))] #average SeaWiFS only
files.chl1.avw=files.chl1[grep(files.chl1, pattern=('_AVW-'))] #weighted average merge of multiple satellite data
## Merge the AV and AVW chl1 product filenames
test=data.frame(files.chl1.ave,stringsAsFactors = FALSE); colnames(test)='chl1'
test2=data.frame(files.chl1.avw,stringsAsFactors = FALSE); colnames(test2)='chl1'
files.chl1.av=rbind(test, test2); rm(test); rm(test2)
## Select just GSM chl1 product filenames
files.chl1.gsm=data.frame(files.chl1[grep(files.chl1, pattern=('_GSM-'))],stringsAsFactors = FALSE);colnames(files.chl1.gsm)='gsm' #GSM merge of multiple satellite data
## HERMES chl2 product files (coastal, limited data)
files.chl2=list.files(wd, pattern=('_CHL2_8D')) #HERMES chl2, only for MER and OLA sats (limited data)
files.chl2=data.frame(files.chl2, stringsAsFactors = FALSE);colnames(files.chl2)='chl2'
## HERMES OC5 product filenames
files.oc5=list.files(wd, pattern=('CHL-OC5_')) #HERMES oc5 algorithm for coastal waters
files.oc5=data.frame(files.oc5, stringsAsFactors = F);colnames(files.oc5)='oc5' #HERMES oc5 algorithm for coastal waters
### sort data lists to make sure it is in chronological order
av.files=sort(files.chl1.av[,1])
gsm.files=sort(files.chl1.gsm[,1])
oc5.files=sort(files.oc5[,1])
chl2.files=sort(files.chl2[,1])
#get dates
av.dates.8d=list()
for (i in 1:length(av.files)){
av.dates.8d[[i]]=strsplit(av.files[i],split="_", fixed=TRUE)[[1]][2]
}
test=do.call(rbind, strsplit(unlist(av.dates.8d), split="-", fixed=T))
dates.8d=matrix(NA, nrow=length(test[,1]), ncol=3)
dates.8d=data.frame(dates.8d)
dates.8d$X1=as.numeric(substr(test[,1], 1,4)) #YYYY
dates.8d$X2=substr(test[,1], 5,6) #MM
dates.8d$X3=substr(test[,1], 7,8) #DD
dates.8d$Y1=substr(test[,2], 1,4) #YYYY
dates.8d$Y2=substr(test[,2], 5,6) #MM
dates.8d$Y3=substr(test[,2], 7,8) #DD
dates.8d$F1=paste(dates.8d[,1], dates.8d[,2], dates.8d[,3], sep='-')
dates.8d$F2=paste(dates.8d[,4], dates.8d[,5], dates.8d[,6], sep='-')
dates.8d$DOY1=as.numeric(strftime(dates.8d$F1, '%j'))
dates.8d$DOY2=as.numeric(strftime(dates.8d$F2, '%j'))
## add week of year (should be 46 per year for 8-day files)
table(dates.8d$X1)
dates.8d$week=c(seq(from=31, to=46, by=1), rep(seq(from=1, to=46, by=1), 21), seq(from=1, to=23, by=1))
### create raster stacks of data
chl.av.8d=nc2raster(files.chl1.av[,1], 'CHL1_mean')
chl.gsm.8d=nc2raster(files.chl1.gsm[,1], 'CHL1_mean')
chl.oc5.8d=nc2raster(files.oc5[,1], 'CHL-OC5_mean')
#testing
plotChlRaster(chl.av.8d, 5, 5, limit=T, av.dates.8d)
### HERMES monthly composites ###
# setwd('G:/1 RM/3 gridded data/HERMES merged CHL 25km')
setwd("C:/Users/ryan.morse/Desktop/Iomega Drive Backup 20171012/1 RM/3 gridded data/HERMES_monthly")
setwd('/media/ryan/Iomega_HDD/1 RM/3 gridded data/HERMES_monthly')
wd=getwd()
## HERMES chl1 product for class 1 ocean waters, inlcudes GSM, AV, and AVW files
files.chl1=list.files(wd, pattern=('_CHL1_MO'))
files.chl1.ave=files.chl1[grep(files.chl1, pattern=('_AV-'))] #average SeaWiFS only
files.chl1.avw=files.chl1[grep(files.chl1, pattern=('_AVW-'))] #weighted average merge of multiple satellite data
## Merge the AV and AVW chl1 product filenames
test=data.frame(files.chl1.ave,stringsAsFactors = FALSE); colnames(test)='chl1'
test2=data.frame(files.chl1.avw,stringsAsFactors = FALSE); colnames(test2)='chl1'
files.chl1.av=rbind(test, test2); rm(test); rm(test2)
## Select just GSM chl1 product filenames
files.chl1.gsm=data.frame(files.chl1[grep(files.chl1, pattern=('_GSM-'))],stringsAsFactors = FALSE);colnames(files.chl1.gsm)='gsm' #GSM merge of multiple satellite data
## HERMES chl2 product files (coastal, limited data)
files.chl2=list.files(wd, pattern=('_CHL2_MO')) #HERMES chl2, only for MER and OLA sats (limited data)
files.chl2=data.frame(files.chl2, stringsAsFactors = FALSE);colnames(files.chl2)='chl2'
## HERMES OC5 product filenames
files.oc5=list.files(wd, pattern=('CHL-OC5_')) #HERMES oc5 algorithm for coastal waters
files.oc5=data.frame(files.oc5, stringsAsFactors = F);colnames(files.oc5)='oc5' #HERMES oc5 algorithm for coastal waters
### sort data lists to make sure it is in chronological order
av.files=sort(files.chl1.av[,1])
gsm.files=sort(files.chl1.gsm[,1])
oc5.files=sort(files.oc5[,1])
chl2.files=sort(files.chl2[,1])
#get dates
av.dates=list()
for (i in 1:length(av.files)){
av.dates[[i]]=strsplit(av.files[i],split="_", fixed=TRUE)[[1]][2]
}
# nc1=nc_open(av.files[1])
# lon.av=ncvar_get(nc1, 'lon')
# lat.av=ncvar_get(nc1, 'lat')
# chl.av.1=ncvar_get(nc1, 'CHL1_mean')
# time.occi=ncvar_get(nc1, 'time') #days since Jan 1, 1970
# dim(chl.occi)
### create raster stacks of data
chl.av=nc2raster(files.chl1.av[,1], 'CHL1_mean')
chl.gsm=nc2raster(files.chl1.gsm[,1], 'CHL1_mean')
chl.oc5=nc2raster(files.oc5[,1], 'CHL-OC5_mean')
nc1=nc_open(files.oc5[1,1])
i=20
plot(chl.av[[i]], main=test1[i])
# testing...
plotChlRaster(chl.gsm, 5, 5, limit=T, av.dates)
plotChlRaster(chl.av, 5, 5, limit=T, av.dates)
plotChlRaster(chl.av, 5, 20, limit=F, av.dates)
# create time series by box for ecomon strata
m.av=extract_calc(chl.av[[1:262]], ecomon.strata)
m.gsm=extract_calc(chl.gsm[[1:262]], ecomon.strata)
m.oc5=extract_calc(chl.oc5[[1:262]], ecomon.strata)
#CBay
plot(m.av[6,], type='l', main='Box 6 CBay')
lines(m.gsm[6,], col='red')
#GBK
plot(m.av[30,], type='l', main='Box 30 GBK')
lines(m.gsm[30,], col='red')
#NY
plot(m.av[17,], type='l', main='Box 17 Hudson River')
lines(m.gsm[17,], col='red')
plot((m.av[1,]-m.gsm[1,]), type='l')
abline(h=0, lty=2)
#
plot(m.av[26,], type='l', main='Box 26 GBK outer shelf')
lines(m.gsm[26,], col='red')
#
ii=6
plot(m.oc5[ii,], type='l', main=paste('Box',ii))
lines(m.gsm[ii,], col='red')
lines(m.av[ii,], col='blue')
## spring yearly means stacked raster - these are already monthly means, so just take mean for season and stack
mn.lst=c(2,3,4); season='Spring'
files.02GSM=list.files(wd, pattern='0201');files.02GSM=grep(files.02GSM, pattern='_GSM-', inv=F, value=T)
files.02GSM=files.02GSM[c(1:4,6:length(files.02GSM))] #drop 20020131-20020228
files.03GSM=list.files(wd, pattern='0301');files.03GSM=grep(files.03GSM, pattern='_GSM-', inv=F, value=T)
files.03GSM=files.03GSM[c(1:5,7:length(files.03GSM))] #drop 20030131-20030228
files.04GSM=list.files(wd, pattern='0401');files.04GSM=grep(files.04GSM, pattern='_GSM-', inv=F, value=T)
files.04GSM=files.04GSM[c(1:6,8:length(files.04GSM))] #drop 20040131-20040228
r1 <- raster(files.02GSM[1], varname = "CHL1_mean")
r2 <- raster(files.03GSM[1], varname = "CHL1_mean")
r3 <- raster(files.04GSM[1], varname = "CHL1_mean")
s <- stack(r1, r2, r3)
x <- reclassify(s, cbind(0, NA))
# shp.dat= mean(x, na.rm=TRUE) # mean of pixels
# fun=function(x) { (range(x, na.rm=T)[2]-range(x, na.rm=T)[1])+1} # inclusive difference between high and low of range
fun=function(x) { if (is.na(x[1])){ NA } else {(range(x, na.rm=T)[2]-range(x, na.rm=T)[1])+1}}
shp.dat= calc(s, fun)
for (i in 2:length(files.02GSM)){
r1 <- raster(files.02GSM[i], varname = "CHL1_mean")
r2 <- raster(files.03GSM[i], varname = "CHL1_mean")
r3 <- raster(files.04GSM[i], varname = "CHL1_mean")
s <- stack(r1, r2, r3)
x <- reclassify(s, cbind(0, NA))
# shp.dat2=mean(x, na.rm=TRUE) # mean of pixels
shp.dat2= calc(s, fun) # function above (range)
shp.dat=stack(shp.dat, shp.dat2)
}
if (season == 'Spring'){
yrlist2=yrlist[-(1),] # drop 1997 for spring (no data)
}
# if (season == 'Fall'){
# yrlist=yrlist[-(19),] # drop 2015 for fall (data not downloaded yet)
# }
# save(shp.dat, file='Spr_25km_chl_gsm.rdata')
wd3= "G:/1 RM/2 Plankton Spatial Plots/data"
filename=paste(season, datalab, yrlist2[1,1],yrlist[length(yrlist[,1]),1],"shp.dat.rdata", sep="_")
mypath=file.path(wd3, filename)
save(shp.dat, file=mypath)
#
# v2=list()
# for(i in 1:dim(shp.dat)[3]){
# v=extract(shp.dat[[i]], neus.shp)
# v1=lapply(v, function(x) mean(x, na.rm=T))
# v2[i]=list(v1)
# # names(v1[i])=seq(0:29)
# }
# m=matrix(unlist(v2), ncol=18, nrow=30) # box 0-29 =rows, years =cols
#
# v1=extract(shp.dat[[1]], neus.shp)
# v1m=lapply(v1, function(x) mean(x, na.rm=T))
#
files.01GSM=list.files(wd, pattern='0101');files.01GSM=grep(files.01GSM, pattern='_GSM-', inv=F, value=T)
ii=data.frame(files.01GSM)
files.01GSM=files.01GSM[c(1:13,17:22)] #21
ii=data.frame(files.01GSM)
files.02GSM=list.files(wd, pattern='0201');files.02GSM=grep(files.02GSM, pattern='_GSM-', inv=F, value=T)
ii=data.frame(files.02GSM)
files.02GSM=files.02GSM[c(1:4,6:20)] #drop 20020131-20020228
files.03GSM=list.files(wd, pattern='0301');files.03GSM=grep(files.03GSM, pattern='_GSM-', inv=F, value=T)
ii=data.frame(files.03GSM)
files.03GSM=files.03GSM[c(1:5,7:20)] #drop 20030131-20030228
files.04GSM=list.files(wd, pattern='0401');files.04GSM=grep(files.04GSM, pattern='_GSM-', inv=F, value=T)
ii=data.frame(files.04GSM)
files.04GSM=files.04GSM[c(1:6,8:20)] #drop 20040131-20040228
files.05GSM=list.files(wd, pattern='0501');files.05GSM=grep(files.05GSM, pattern='_GSM-', inv=F, value=T)
ii=data.frame(files.05GSM)
files.05GSM=files.05GSM[c(1:7,9:20)]
files.06GSM=list.files(wd, pattern='0601');files.06GSM=grep(files.06GSM, pattern='_GSM-', inv=F, value=T)
ii=data.frame(files.06GSM)
files.06GSM=files.06GSM[c(1:8,10:20)]
files.07GSM=list.files(wd, pattern='0701');files.07GSM=grep(files.07GSM, pattern='_GSM-', inv=F, value=T)
ii=data.frame(files.07GSM)
files.07GSM=files.07GSM[c(1:9,11:20)]
files.08GSM=list.files(wd, pattern='0801');files.08GSM=grep(files.08GSM, pattern='_GSM-', inv=F, value=T)
ii=data.frame(files.08GSM)
files.08GSM=files.08GSM[c(1:10,12:20)]
files.09GSM=list.files(wd, pattern='0901');files.09GSM=grep(files.09GSM, pattern='_GSM-', inv=F, value=T)
ii=data.frame(files.09GSM)
files.09GSM=files.09GSM[c(1:11,13:20)]
files.10GSM=list.files(wd, pattern='1001');files.10GSM=grep(files.10GSM, pattern='_GSM-', inv=F, value=T)
ii=data.frame(files.10GSM)
files.10GSM=files.10GSM[c(1:12,14:20)]
files.11GSM=list.files(wd, pattern='1101');files.11GSM=grep(files.11GSM, pattern='_GSM-', inv=F, value=T)
ii=data.frame(files.11GSM)
files.11GSM=files.11GSM[c(1:13,15:20)]
files.12GSM=list.files(wd, pattern='1201');files.12GSM=grep(files.12GSM, pattern='_GSM-', inv=F, value=T)
ii=data.frame(files.12GSM)
files.12GSM=files.12GSM[c(1:14,16:20)]
# Chl.time=seq(ISOdate(1998,1,15), ISOdate(2016,12,15), "month") # monthly mean values 1998-2016
### read in calibration files from seaBASS
setwd('C:/Users/ryan.morse/Documents/GitHub/JPSS/calibration')
mrs=read.csv('1563559063629053_chlor_a.csv', skip=26, header=T, stringsAsFactors = F); mrs=mrs[-c(1:2),] # MERIS
swf=read.csv('1563558739820325_chlor_a.csv', skip=26, header=T, stringsAsFactors = F); swf=swf[-c(1:2),] # SeaWiFS
mds=read.csv('1563558871350863_chlor_a.csv', skip=26, header=T, stringsAsFactors = F); mds=mds[-c(1:2),] # Modis
vrs=read.csv('1563558999992368_chlor_a.csv', skip=26, header=T, stringsAsFactors = F); vrs=vrs[-c(1:2),] # Viirs-snpp
mrs$longitude=as.numeric(mrs$longitude); mrs$latitude=as.numeric(mrs$latitude); mrs$insitu_chlor_a=as.numeric(mrs$insitu_chlor_a)
coordinates(mrs)=~longitude+latitude#transform to Spatialpointsdataframe
proj4string(mrs)=CRS("+proj=longlat +datum=NAD83 +no_defs +ellps=GRS80 +towgs84=0,0,0") #ensure same projection
pointsin=over(mrs, NES.shp) #find which boxes samples belong to
# pointsin=over(mrs, ecomon.strata)
map("worldHires", xlim=c(-77,-65),ylim=c(35,45), fill=T,border=0,col="gray70")
map.axes(las=1)
points(mrs)
swf$longitude=as.numeric(swf$longitude); swf$latitude=as.numeric(swf$latitude); swf$insitu_chlor_a=as.numeric(swf$insitu_chlor_a)
coordinates(swf)=~longitude+latitude#transform to Spatialpointsdataframe
proj4string(swf)=CRS("+proj=longlat +datum=NAD83 +no_defs +ellps=GRS80 +towgs84=0,0,0") #ensure same projection
pointsin=over(swf, NES.shp) #find which boxes samples belong to
# pointsin=over(swf, ecomon.strata)
map("worldHires", xlim=c(-77,-65),ylim=c(35,45), fill=T,border=0,col="gray70")
map.axes(las=1)
points(swf)
mds$longitude=as.numeric(mds$longitude); mds$latitude=as.numeric(mds$latitude); mds$insitu_chlor_a=as.numeric(mds$insitu_chlor_a)
coordinates(mds)=~longitude+latitude#transform to Spatialpointsdataframe
proj4string(mds)=CRS("+proj=longlat +datum=NAD83 +no_defs +ellps=GRS80 +towgs84=0,0,0") #ensure same projection
pointsin=over(mds, NES.shp) #find which boxes samples belong to
# pointsin=over(mds, ecomon.strata)
map("worldHires", xlim=c(-77,-65),ylim=c(35,45), fill=T,border=0,col="gray70")
map.axes(las=1)
points(mds)
vrs$longitude=as.numeric(vrs$longitude); vrs$latitude=as.numeric(vrs$latitude); vrs$insitu_chlor_a=as.numeric(vrs$insitu_chlor_a)
coordinates(vrs)=~longitude+latitude#transform to Spatialpointsdataframe
proj4string(vrs)=CRS("+proj=longlat +datum=NAD83 +no_defs +ellps=GRS80 +towgs84=0,0,0") #ensure same projection
pointsin=over(vrs, NES.shp) #find which boxes samples belong to
# pointsin=over(vrs, ecomon.strata)
map("worldHires", xlim=c(-77,-65),ylim=c(35,45), fill=T,border=0,col="gray70")
map.axes(las=1)
points(vrs)
### get calibration Chl from WOD
d2='C:/Users/ryan.morse/Documents/GitHub/JPSS/calibration/WOD'
d2='~/Git/JPSS/calibration/WOD'
d2='/home/ryan/Git/JPSS_2/JPSS/calibration/WOD'
ncfiles=list.files(path=d2, pattern='.nc')
nc.str=strsplit(ncfiles, '.nc')
i=1 #CTD hi res
nc1=nc_open(paste(d2, '/', ncfiles[i], sep=''))
# nc1=nc_open(ncfiles[i])
wod.chl=ncvar_get(nc1, 'Chlorophyll')
wod.lat=ncvar_get(nc1, 'lat')
wod.lon=ncvar_get(nc1, 'lon')
wod.time=ncvar_get(nc1, 'time') #days since Jan 1, 1770
wod.z=ncvar_get(nc1, 'z')
nc1$var$time$units
# i=4 #profiler buoy samples
# nc1=nc_open(ncfiles[i])
# wod.pfl.chl=ncvar_get(nc1, 'Chlorophyll')
# wod.pfl.lat=ncvar_get(nc1, 'lat')
# wod.pfl.lon=ncvar_get(nc1, 'lon')
# wod.pfl.time=ncvar_get(nc1, 'time') #days since Jan 1, 1770
# i=2 #glider samples
# nc1=nc_open(ncfiles[i])
# wod.gld.chl=ncvar_get(nc1, 'Chlorophyll')
# wod.gld.lat=ncvar_get(nc1, 'lat')
# wod.gld.lon=ncvar_get(nc1, 'lon')
# wod.gld.time=ncvar_get(nc1, 'time') #days since Jan 1, 1770
# i=3 #bottle samples
# nc1=nc_open(ncfiles[i])
# wod.osd.chl=ncvar_get(nc1, 'Chlorophyll')
# wod.osd.lat=ncvar_get(nc1, 'lat')
# wod.osd.lon=ncvar_get(nc1, 'lon')
# wod.osd.time=ncvar_get(nc1, 'time') #days since Jan 1, 1770
## WOD CTD samples, surface depth
wod.chl.df=data.frame(wod.chl[which(wod.z==0)], wod.lon, wod.lat, wod.z[which(wod.z==0)], wod.time)
test=month.day.year(wod.chl.df$wod.time, c(1,1,1770))
wod.chl.df$month=test$month
wod.chl.df$day=test$day
wod.chl.df$year=test$year
wod.chl.df2=wod.chl.df[which(wod.chl.df$year>1996),]
colnames(wod.chl.df2)=c('chl', 'lon', 'lat', 'z', 'jday', 'month', 'day', 'year')
wod.chl.df2$F1=paste(wod.chl.df2$year, wod.chl.df2$month, wod.chl.df2$day, sep='-')
wod.chl.df2$DOY=as.numeric(strftime(wod.chl.df2$F1, '%j'))
wod.chl.df2=wod.chl.df2[complete.cases(wod.chl.df2),]
save(wod.chl.df2, file='/home/ryan/Git/JPSS_2/JPSS/WOD_CTD_surface_chl.RData')
barplot(table(round(wod.chl.df2$chl, digits=1)))
barplot(table(wod.chl.df2$month))
barplot(table(wod.chl.df2$year))
# coordinates(wod.chl.df2)=~lon+lat #transform to Spatialpointsdataframe
# proj4string(wod.chl.df2)=CRS("+proj=longlat +datum=NAD83 +no_defs +ellps=GRS80 +towgs84=0,0,0") #ensure same projection
# pointsin=over(wod.chl.df2, NES.shp) #find which boxes samples belong to
map("worldHires", xlim=c(-77,-65),ylim=c(35,45), fill=T,border=0,col="gray70")
map.axes(las=1)
points(wod.chl.df2$lon, wod.chl.df2$lat)
points(wod.chl.df2)
### build initial list of time matchups - indicates the dimension of the chl raster stack to extract data from
# yy=unique(wod.chl.df2$year) # unique years
wod.chl.df2$smatch=NA
# wod.chl.df2$DOYmed=NA
wod.chl.df2$sDOY1=NA
wod.chl.df2$sDOY2=NA
for(i in 1:length(wod.chl.df2$lon)){
ylim=which(dates.8d$X1==wod.chl.df2$year[i])
xmn=which(dates.8d$DOY1[ylim]<=wod.chl.df2$DOY[i])#[dates.8d$Y1==yj]
xmx=which(dates.8d$DOY2[ylim]>=wod.chl.df2$DOY[i])#[dates.8d$Y1==yj]
both=ylim[which(xmn%in%xmx)]
if(length(both)<1){
next
}
else {
wod.chl.df2$smatch[i]=both
wod.chl.df2$sDOY1[i]=dates.8d$DOY1[both]
wod.chl.df2$sDOY2[i]=dates.8d$DOY2[both]
}
}
wod.chl.df2$DOYmed=round((wod.chl.df2$sDOY1+wod.chl.df2$sDOY2)/2, digits=0) # median satellite DOY
wod.chl.df2$ddif=wod.chl.df2$DOY-wod.chl.df2$DOYmed # difference from median satellite date
### now extract from raster
# i=2746
# tt=extract(chl.gsm.8d[[140]], wod.chl.df2[i,], method='bilinear', fun='mean', na.rm=T)
wod.chl.df2=wod.chl.df2[complete.cases(wod.chl.df2$smatch),]
WOD=wod.chl.df2
wod.chl.df2$gsm=NA
wod.chl.df2$oc5=NA
wod.chl.df2$av=NA
wod.chl.df2$occi=NA
wod.chl.df2$occiv5=NA
coordinates(wod.chl.df2)=~lon+lat #transform to Spatialpointsdataframe
### method=bilinear #interpolates value from 4 nearest raster cells #simple is for cell only
for(i in 1:length(wod.chl.df2$chl)){
wod.chl.df2$gsm[i]=raster::extract(chl.gsm.8d[[wod.chl.df2$smatch[i]]], wod.chl.df2[i,], method='bilinear', fun='mean', na.rm=T)
# wod.chl.df2$av[i]=extract(chl.av.8d[[wod.chl.df2$smatch[i]]], wod.chl.df2[i,], method='bilinear', fun='mean', na.rm=T)
# wod.chl.df2$oc5[i]=extract(chl.oc5.8d[[wod.chl.df2$smatch[i]]], wod.chl.df2[i,], method='bilinear', fun='mean', na.rm=T)
wod.chl.df2$occi[i]=raster::extract(occi[[wod.chl.df2$smatch[i]]], wod.chl.df2[i,], method='bilinear', fun='mean', na.rm=T)
wod.chl.df2$occiv5[i]=raster::extract(occi.v5[[wod.chl.df2$smatch[i]]], wod.chl.df2[i,], method='bilinear', fun='mean', na.rm=T)
if (i%%100==0){
}
}
## add time to dataframe for satellite matchup
wod.chl.df2$jtime=wod.chl.df2$jday-floor(wod.chl.df2$jday)
wod.chl.df2$time=format(times(wod.chl.df2$jtime))
write.csv(wod.chl.df2, file='WOD_calibration_to_surface_chlorophyll_8day.csv')
write.csv(wod.chl.df2, file='WOD_surf.csv')
WOD$gsm=wod.chl.df2$gsm
WOD$av=wod.chl.df2$av
WOD$oc5=wod.chl.df2$oc5
WOD=WOD[complete.cases(WOD$gsm),]
WOD=WOD[complete.cases(WOD$chl),]
colorpal=viridis::viridis(8)
# plot(log10(WOD$chl)~log10(WOD$gsm), type='n')#, color=colorpal[WOD$ddif+4])
# # points(log10(WOD$chl),log10(WOD$gsm), type='p', col=colorpal[WOD$ddif+4])
# points(log10(WOD$gsm), log10(WOD$chl), type='p', col=colorpal[WOD$ddif+4])
# abline(0,1)
# Non log transformed data
plot(WOD$chl~WOD$oc5, type='n')#, color=colorpal[WOD$ddif+4])
points(WOD$oc5, WOD$chl, type='p', col=colorpal[WOD$ddif+4])
abline(0,1)
reg1=lm(WOD$chl~WOD$oc5)
summary(reg1)
abline(reg1$coefficients[1], reg1$coefficients[2], col='red')
# plot(x=WOD$gsm, y=WOD$chl, log='xy') #, color=colorpal[WOD$ddif+4])
# abline(0,1)
# abline(reg1$coefficients[1], reg1$coefficients[2], col='blue')
# barplot(table(WOD$ddif))
## log transform both X and Y
x=log10(WOD$gsm+0.001)
y=log10(WOD$chl+0.001)
reg1=lm(y~x)
summary(reg1)
xy=data.frame(x,y)
xy=xy[complete.cases(x),]
plot(y~x, type='n')#, log='xy')#, color=colorpal[WOD$ddif+4])
points(log10(WOD$gsm), log10(WOD$chl), type='p', col=colorpal[WOD$ddif+4])
abline(0,1)
abline(reg1$coefficients[1], reg1$coefficients[2], col='red')
library(plotrix)
taylor.diagram(wod.chl.df2$chl, wod.chl.df2$oc5, col='red')
taylor.diagram(wod.chl.df2$chl, wod.chl.df2$gsm, add=T, col='blue')
taylor.diagram(wod.chl.df2$chl, wod.chl.df2$av, add=T, col='green')
taylor.diagram(wod.chl.df2$chl, wod.chl.df2$occi, add=T, col='black')
taylor.diagram(wod.chl.df2$chl, wod.chl.df2$occiv5, add=T, col='red')
#subset to low or high chl
test=wod.chl.df2[which(wod.chl.df2$chl<1),]
taylor.diagram(test$chl, test$oc5, col='red')
taylor.diagram(test$chl, test$gsm, add=T, col='blue')
taylor.diagram(test$chl, test$av, add=T, col='green')
taylor.diagram(test$chl, test$occi, add=T, col='black')
taylor.diagram(test$chl, test$occiv5, add=T, col='red')
RMSE = function(m, o){
sqrt(mean((m - o)^2))
}
# Function that returns Mean Absolute Error
mae <- function(error)
{
mean(abs(error))
}
# Function that returns Root Mean Squared Error
rmse <- function(error)
{
sqrt(mean(error^2))
}
# Calculate error
error <- actual - predicted
test=WOD[complete.cases(WOD$occciv5),]
RMSE(o=test$chl, m=test$occciv5)
dim(test)
test=WOD[complete.cases(WOD$occciv4),]
RMSE(o=test$chl, m=test$occciv4)
dim(test)
test=WOD[complete.cases(WOD$gsm),]
RMSE(o=test$chl, m=test$gsm)
dim(test)
test=WOD[complete.cases(WOD),] # all vars, equal lenght (much data lost, all remaining in NES)
RMSE(o=test$chl, m=test$occciv5)
RMSE(o=test$chl, m=test$occciv4)
RMSE(o=test$chl, m=test$gsm)
### look at residuals
test$rv4=test$chl-test$occi
test$rv5=test$chl-test$occiv5
test$rgsm=test$chl-test$gsm
mae(test$rgsm)
mae(test$rv4)
mae(test$rv5)
rmse(test$rgsm)
rmse(test$rv4)
rmse(test$rv5)
plot(test$rv4,type='p', ylim=c(-10, 10), main='v4')
plot(test$rv5,type='p', ylim=c(-10, 10), main='v5')
plot(test$rgsm,type='p', ylim=c(-10, 10), main='gsm')
plot(test$rv4~test$occciv4, type='p', main='v4 resid', ylim=c(-5, 15), xlim=c(0, 8))
plot(test$rv5~test$occciv5, type='p', main='v5 resid', ylim=c(-5, 15), xlim=c(0, 8))
plot(test$rgsm~test$gsm, type='p', main='gsm resid', ylim=c(-5, 15), xlim=c(0, 8))
## log scale
plot(test$occciv4 ~ test$chl,type='p', main='v4', log='xy'); abline(a=0, b=1)
plot(test$occciv5 ~ test$chl,type='p', main='v5', log='xy'); abline(a=0, b=1)
plot(test$gsm ~ test$chl,type='p', main='gsm', log='xy'); abline(a=0, b=1)
plot(log(test$occciv4+1) ~ log(test$chl+1) ,type='p', ylim=c(0, 2), xlim=c(0,2), main='v4'); abline(a=0, b=1)
plot(log(test$occciv5+1) ~log(test$chl+1),type='p', ylim=c(0, 2), xlim=c(0,2), main='v5'); abline(a=0, b=1)
plot(log(test$gsm+1) ~log(test$chl+1) ,type='p', ylim=c(0, 2), xlim=c(0,2), main='gsm'); abline(a=0, b=1)
### read in in situ chlorophyll-satellite matchups from NASA SEABASS
sbmodisa=read.csv('/home/ryan/Downloads/1612220222584465_chlor_a.csv', header = F, stringsAsFactors = F, skip = 29)
t=readLines('/home/ryan/Downloads/1612220222584465_chlor_a.csv', n=28)
t2=strsplit(t[27], split=',')
colnames(sbmodisa)=unlist(t2)
sbseawifs=read.csv('/home/ryan/Downloads/161222030910903_chlor_a.csv', header = F, stringsAsFactors = F, skip = 29)
t=readLines('/home/ryan/Downloads/161222030910903_chlor_a.csv', n=28)
t2=strsplit(t[27], split=',')
colnames(sbseawifs)=unlist(t2)
sbmeris=read.csv('/home/ryan/Downloads/1612220388699824_chlor_a.csv', header = F, stringsAsFactors = F, skip = 29)
t=readLines('/home/ryan/Downloads/1612220388699824_chlor_a.csv', n=28)
t2=strsplit(t[27], split=',')
colnames(sbmeris)=unlist(t2)
sbviirssnpp=read.csv('/home/ryan/Downloads/1612220443903143_chlor_a.csv', header = F, stringsAsFactors = F, skip = 29)
t=readLines('/home/ryan/Downloads/1612220443903143_chlor_a.csv', n=28)
t2=strsplit(t[27], split=',')
colnames(sbviirssnpp)=unlist(t2)
sbmodist=read.csv('/home/ryan/Downloads/1612220548231565_chlor_a.csv', header = F, stringsAsFactors = F, skip = 29)
t=readLines('/home/ryan/Downloads/1612220548231565_chlor_a.csv', n=28)
t2=strsplit(t[27], split=',')
colnames(sbmodist)=unlist(t2)
tr1=sbmodisa %>% select( latitude, longitude, date_time, insitu_chlor_a, aqua_chlor_a)
tr2=sbmodist %>% select( latitude, longitude, date_time, insitu_chlor_a, terra_chlor_a)
tr3=sbmeris %>% select( latitude, longitude, date_time, insitu_chlor_a, meris_chlor_a)
tr4=sbseawifs %>% select( latitude, longitude, date_time, insitu_chlor_a, seawifs_chlor_a)
tr5=sbviirssnpp %>% select( latitude, longitude, date_time, insitu_chlor_a, viirs_chlor_a)
# tr=tr1 %>% left_join(tr2, by=c('latitude', 'longitude', 'date_time'))
tr=tr1 %>% full_join(tr2, by=c('latitude', 'longitude', 'date_time'), name=NULL)
tr=tr%>% full_join(tr3, by=c('latitude', 'longitude', 'date_time'), name=NULL)
tr=tr%>% full_join(tr4, by=c('latitude', 'longitude', 'date_time'), name=NULL)
tr=tr%>% full_join(tr5, by=c('latitude', 'longitude', 'date_time'), name=NULL)
### merge all in-situ chlorophylls together, drop merged repeats
insitu_chl=tr %>% select(insitu_chlor_a, insitu_chlor_a.x, insitu_chlor_a.y, insitu_chlor_a.x.x, insitu_chlor_a.y.y) %>%
mutate(ischl=rowMeans(., na.rm=T))
tr$insitu_chl_final=insitu_chl$ischl
tr=tr %>% select(-insitu_chlor_a, -insitu_chlor_a.x, -insitu_chlor_a.y, -insitu_chlor_a.x.x, -insitu_chlor_a.y.y)
### read in satellite and ship matchups from Kim (daily)
satshpv4=read.csv('/home/ryan/Downloads/SATSHIP_L2-JPSS_NEC_SEABASS-CHLOR_A-OCI-OCCCI.CSV', sep = ',',
header = F, stringsAsFactors = F, skip=1)
t=readLines('/home/ryan/Downloads/SATSHIP_L2-JPSS_NEC_SEABASS-CHLOR_A-OCI-OCCCI.CSV', n=1)
t2=strsplit(t, split=',')
colnames(satshpv4)=unlist(t2)
satshpv4$SHIP_DATE=format(satshpv4$SHIP_DATE, scientific = F)
satshpv4$SHIP_DATE=ymd_hms(satshpv4$SHIP_DATE)
satshpv5=read.csv('/home/ryan/Downloads/SATSHIP_L2-JPSS_NEC_SEABASS-CHLOR_A-CCI-OCCCI_V5.0.CSV',
sep = ',', header=F, stringsAsFactors = F, skip=1)
t=readLines('/home/ryan/Downloads/SATSHIP_L2-JPSS_NEC_SEABASS-CHLOR_A-CCI-OCCCI_V5.0.CSV', n=1)
t2=strsplit(t, split=',')
colnames(satshpv5)=unlist(t2)
satshpv5$SHIP_DATE=format(satshpv5$SHIP_DATE, scientific = F)
satshpv5$SHIP_DATE=ymd_hms(satshpv5$SHIP_DATE)
## separate out 3x3 pixels around center chl, take geometric mean (or median)
### For OCCCI v5 data extractions
satv5chl=satshpv5 %>% tidyr::separate(SAT_CHLOR_A_CCI, sep=";", into=paste("v", 1:9, sep=''), convert=T) %>% select(v1:v9)
# satv5chl$gm=apply(satv5chl[,1:9], 1, FUN=gm_mean, na.rm=T)
satv5chl$med=apply(satv5chl[,1:9], 1, median, na.rm=T)
satshpv5$SAT_MED_CHLOR=satv5chl$med
tv5=satshpv5[complete.cases(satshpv5$SAT_MED_CHLOR),]
### now do for v4 OCCCI data
satv4chl=satshpv4 %>% tidyr::separate(SAT_CHLOR_A_OCI, sep=";", into=paste("v", 1:9, sep=''), convert=T) %>% select(v1:v9)
satv4chl$med=apply(satv4chl[,1:9], 1, median, na.rm=T)
satshpv4$SAT_MED_CHLOR=satv4chl$med
tv4=satshpv4[complete.cases(satshpv4$SAT_MED_CHLOR),]
### read sat-ship matchup file from Kyle turner
ktsatmat=readxl::read_excel('/home/ryan/Downloads/nes_insitu_satellite_chl_upper10m.xlsx', na="NaN",
col_types=c('numeric', 'numeric', 'guess', 'numeric', 'text', 'numeric','numeric','numeric','numeric','numeric','numeric','numeric','numeric'))
# ktsatmat=ktsatmat %>% mutate_at(vars(`viirs_snpp_chlor_a [mg m-3]`, `snpp_time_dif [hours]`, `viirs_noaa20_chlor_a [mg m-3]`, `noaa20_time_dif [hours]`,
# `modis_aqua_chlor_a [mg m-3]`, `modis_time_dif [hours]`, `occci_chlor_a [mg m-3]`), na_if, "NaN")
# ktsatmat[ktsatmat=="NaN"]=NA #as.numeric(ktsatmat)
t=ktsatmat[complete.cases(ktsatmat$`occci_chlor_a [mg m-3]`),]
plot(log(t$`in_situ_chl [mg m-3]`)~log(t$`occci_chlor_a [mg m-3]`), type='p')
abline(a=0, b=1)
RMSE(o=t$`in_situ_chl [mg m-3]` , m=t$`occci_chlor_a [mg m-3]`)
t2=satshpv5 %>% filter(CRUISE!='SeaBASS')
t3=satshpv5 %>% filter(CRUISE=='SeaBASS')
table(year(t3$SHIP_DATE))
### Turner merge v5 sat-matchups from Kim
ktest5=left_join(tv5,ktsatmat, by=c("date"="datetime", "SHIP_LAT"="lat", "SHIP_LON"="lon"))
ktest50=ktest5[complete.cases(ktest5$`in_situ_chl [mg m-3]`),]
plot(log(ktest50$`in_situ_chl [mg m-3]`) ~log(ktest50$SAT_MED_CHLOR), type='p', main="OCCCI v5.0 median vs in situ chl", xlim=c(-3,3), ylim=c(-4,4))
abline(a=0, b=1)
a=RMSE(o=log(ktest50$`in_situ_chl [mg m-3]`) , m=log(ktest50$SAT_MED_CHLOR))
text(-2,3, paste('RMSE: ', round(a, digits=2),sep=''))
text(-2,2, paste('N: ', dim(ktest50)[1],sep=''))
### Turner merge v4 sat-matchups from Kim
ktest4=left_join(tv4,ktsatmat, by=c("date"="datetime", "SHIP_LAT"="lat", "SHIP_LON"="lon"))
ktest40=ktest4[complete.cases(ktest4$`in_situ_chl [mg m-3]`),]
plot(log(ktest40$`in_situ_chl [mg m-3]`) ~log(ktest40$SAT_MED_CHLOR), type='p', main="OCCCI v4.2 median vs in situ chl", xlim=c(-3,3), ylim=c(-4,4))
abline(a=0, b=1)
a=RMSE(o=log(ktest40$`in_situ_chl [mg m-3]`) , m=log(ktest40$SAT_MED_CHLOR))
text(-2,3, paste('RMSE: ', round(a, digits=2),sep=''))
text(-2,2, paste('N: ', dim(ktest40)[1],sep=''))
### now add in tr to see if any additional stations are added... YES
test=full_join(ktest5,tr, by=c("date", "SHIP_LAT"="latitude", "SHIP_LON"="longitude"))
test$insituchl=apply(test[,c(29,43)], 1, mean, na.rm=T)
# sum(is.na(test$`in_situ_chl [mg m-3]`)) #32637
# sum(is.na(test$insitu_chl_final)) #32854
# sum(is.na(test$insituchl)) #32450
test5=test[complete.cases(test$insituchl),] # 916 using median of 3x3 pixels (na.rm=T)
plot(log(test5$insituchl) ~log(test5$SAT_MED_CHLOR), type='p', main="OCCCI v5.0 median vs in situ chl", xlim=c(-3,3), ylim=c(-4,4))
abline(a=0, b=1)
a=RMSE(o=log(test5$insituchl) , m=log(test5$SAT_MED_CHLOR))
text(-2,3, paste('RMSE: ', round(a, digits=2),sep=''))
text(-2,2, paste('N: ', dim(test5)[1],sep=''))
test=left_join(ktest4,tr, by=c("date", "SHIP_LAT"="latitude", "SHIP_LON"="longitude"))
test$insituchl=apply(test[,c(28,42)], 1, mean, na.rm=T)
sum(is.na(test$`in_situ_chl [mg m-3]`)) #32637
sum(is.na(test$insitu_chl_final)) #32854
sum(is.na(test$insituchl)) #32450
test5=test[complete.cases(test$insituchl),] # 916 using median of 3x3 pixels (na.rm=T)
plot(log(test5$insituchl) ~log(test5$SAT_MED_CHLOR), type='p', main="OCCCI v5.0 median vs in situ chl", xlim=c(-3,3), ylim=c(-4,4))
abline(a=0, b=1)
a=RMSE(o=log(test5$insituchl) , m=log(test5$SAT_MED_CHLOR))
text(-2,3, paste('RMSE: ', round(a, digits=2),sep=''))
text(-2,2, paste('N: ', dim(test5)[1],sep=''))
### compare time stamps -> need to format for comparison
tr$date_time[1] %in% tv5$SHIP_DATE # not formatted properly
a=tr$date_time[1] #"2009-10-19 15:25:00"
b=satshpv5$SHIP_DATE[40149] #"2009-10-19 15:25:00 UTC"
as.Date(lubridate::intersect(a,b), origin = "1970-01-01") #still not formatted properly
as.Date(lubridate::intersect(as.Date(a),as.Date(b)), origin = "1970-01-01") # works
as.Date(tr$date_time[1]) %in% as.Date(tv5$SHIP_DATE) # works
test=as.Date(tr$date_time) %in% as.Date(tv5$SHIP_DATE); sum(test) # works, number of in situ samples in OCCCI v5 database from Kim
test4=as.Date(tr$date_time) %in% as.Date(tv4$SHIP_DATE); sum(test)
### need POSIXct for times to be recognized
as.POSIXct(tr$date_time[1]) %in% as.POSIXct(tv5$SHIP_DATE) # does not work
as.POSIXct(tr$date_time[1], tz="UTC") %in% as.POSIXct(tv5$SHIP_DATE, tz="UTC") # works
### merge on dates ---> DOES NOT WORK
tr$date=as.Date(tr$date_time, origin = "1970-01-01")
tv5$date=as.Date(tv5$SHIP_DATE, origin = "1970-01-01")
test=left_join(tr, tv5, by="date") # WAY TOO LONG
### USING POSIXct with times ---> WORKS
tr$date=as.POSIXct(tr$date_time, tz="UTC") #origin = "1970-01-01")
tv5$date=as.POSIXct(tv5$SHIP_DATE, tz="UTC") #origin = "1970-01-01")
tv4$date=as.POSIXct(tv4$SHIP_DATE, tz="UTC") #origin = "1970-01-01")
### NOW do merge:
### v5 data using median of 3x3 pixels
test=left_join(tr, tv5, by=c("date", "latitude"="SHIP_LAT", "longitude"="SHIP_LON")) # 2605 long
test5=test[complete.cases(test$SAT_MED_CHLOR),] # 916 using median of 3x3 pixels (na.rm=T)
plot(log(test5$insitu_chl_final) ~log(test5$SAT_MED_CHLOR), type='p', main="OCCCI v5.0 median vs in situ chl", xlim=c(-3,3), ylim=c(-4,4))
abline(a=0, b=1)
a=RMSE(o=log(test5$insitu_chl_final) , m=log(test5$SAT_MED_CHLOR))
text(-2,3, paste('RMSE: ', round(a, digits=2),sep=''))
text(-2,2, paste('N: ', dim(test5)[1],sep=''))
### v4 data
test=left_join(tr, tv4, by=c("date", "latitude"="SHIP_LAT", "longitude"="SHIP_LON")) # 2609 long
test4=test[complete.cases(test$SAT_MED_CHLOR),] # 917 using median of 3x3 pixels (na.rm=T)
plot(log(test4$insitu_chl_final) ~log(test4$SAT_MED_CHLOR), type='p', main="OCCCI v4.2 median vs in situ chl",xlim=c(-3,3), ylim=c(-4,4))
abline(a=0, b=1)
a=RMSE(o=log(test4$insitu_chl_final) , m=log(test4$SAT_MED_CHLOR))
text(-2,3, paste('RMSE: ', round(a, digits=2),sep=''))
text(-2,2, paste('N: ', dim(test4)[1],sep=''))
### plot non-log transformed
plot(test4$SAT_MED_CHLOR ~ test4$insitu_chl_final, type='p', main="OCCCI v4.2 median vs in situ chl", ylim=c(0,10), xlim=c(0,20)); abline(a=0, b=1)
a=RMSE(o=test4$insitu_chl_final , m=test4$SAT_MED_CHLOR)
text(18,3, paste('RMSE: ', round(a, digits=2),sep=''))
text(18,2, paste('N: ', dim(test4)[1],sep=''))
plot(test5$SAT_MED_CHLOR ~ test5$insitu_chl_final, type='p', main="OCCCI v5.0 median vs in situ chl", ylim=c(0,10), xlim=c(0,20)); abline(a=0, b=1)
a=RMSE(o=test5$insitu_chl_final , m=test5$SAT_MED_CHLOR)
text(18,3, paste('RMSE: ', round(a, digits=2),sep=''))
text(18,2, paste('N: ', dim(test5)[1],sep=''))
### Plot satellite center values log and non log-transformed
## clean Inf values...
test44=test4
test44$SAT_CENTER_CHLOR_A_OCI[which(is.infinite(test44$SAT_CENTER_CHLOR_A_OCI))]=NA
test44=test44[complete.cases(test44$SAT_CENTER_CHLOR_A_OCI),]
test55=test5
test55$SAT_CENTER_CHLOR_A_CCI[which(is.infinite(test55$SAT_CENTER_CHLOR_A_CCI))]=NA
test55=test55[complete.cases(test55$SAT_CENTER_CHLOR_A_CCI),]
plot(test44$SAT_CENTER_CHLOR_A_OCI ~ test44$insitu_chl_final, type='p', main="OCCCI v4.2 center vs in situ chl", ylim=c(0,10), xlim=c(0,20)); abline(a=0, b=1)
a=RMSE(o=test44$insitu_chl_final , m=test44$SAT_CENTER_CHLOR_A_OCI)
text(18,3, paste('RMSE: ', round(a, digits=2),sep=''))
text(18,2, paste('N: ', dim(test44)[1],sep=''))
plot(test55$SAT_CENTER_CHLOR_A_CCI ~ test55$insitu_chl_final, type='p', main="OCCCI v5.0 center vs in situ chl", ylim=c(0,10), xlim=c(0,20)); abline(a=0, b=1)
a=RMSE(o=test55$insitu_chl_final , m=test55$SAT_CENTER_CHLOR_A_CCI)
text(18,3, paste('RMSE: ', round(a, digits=2),sep=''))
text(18,2, paste('N: ', dim(test55)[1],sep=''))
## log transformed
plot(log(test44$insitu_chl_final) ~log(test44$SAT_CENTER_CHLOR_A_OCI), type='p', main="OCCCI v4.2 center vs in situ chl",xlim=c(-3,3), ylim=c(-4,4))
abline(a=0, b=1)
a=RMSE(o=log(test44$insitu_chl_final) , m=log(test44$SAT_CENTER_CHLOR_A_OCI))
text(-2,3, paste('RMSE: ', round(a, digits=2),sep=''))
text(-2,2, paste('N: ', dim(test44)[1],sep=''))
plot(log(test55$insitu_chl_final) ~log(test55$SAT_CENTER_CHLOR_A_CCI), type='p', main="OCCCI v5 center vs in situ chl",xlim=c(-3,3), ylim=c(-4,4))
abline(a=0, b=1)
a=RMSE(o=log(test55$insitu_chl_final) , m=log(test55$SAT_CENTER_CHLOR_A_CCI))