So far, the rsat packaged has proved most reliable. Better at
accessing, scanning and downloading, though it doesnt include a quick
download option and tends to involve longer processing and coding time.
In-built vignettes seem to be best learning resources
(i.e. rsat::rsat1_search).
In the following chunk, we attempt a sort of ‘sign-in’ blitz. This is a
crude attempt to solve ongoing connection issues between rsat
functions and their registered APIs. Two NASA APIs (USGS-eros &
Earthdata) reported in package documentation. However, source code shows
edits in package updates and github commits. Though, the main issue
seems to be the constant maintenance and reformatting of API end-points
by NASA & USGS. See below some details on connection tests, make sure to
change your username and passwords where suggested. More on GDAL
configuration for passive computing in
gdal_cubes.md
library(rsat)
# rsat::set_credentials("[email protected]", "password") working username!
library(espa.tools)
espa.tools::espa_inventory_login(
usgs_eros_username,
usgs_eros_password,
verbose = F)
library(earthdatalogin)
edl_netrc(
username = "username_here",
password = "password_here",
cloud_config = T # see gdal_cubes.md for GDAL config of cloud computing/links
)Most reliable function from getSpatialData that promtps a PAP sign-in
method:
library(getSpatialData) ## is working ???
getSpatialData::login_earthdata(verbose=T)
getSpatialData::login_USGS(verbose=T)
services()library(sf)
library(tmap)
# search paramters
ip <- st_sf(st_as_sfc(st_bbox(c(
xmin = 35.09583,
xmax = 36.29583,
ymin = -15.90890,
ymax = -14.77500
), crs = 4326)))
toi_2018 <- seq(as.Date("2018-10-01"),as.Date("2021-12-30"),1)
# scan all collections
rsat_products()
# confirm available data of study site/window
browse = rsat_search(
region = ip,
dates = toi_2018,
product = c(
"landsat_ot_c2_l1",
"landsat_etm_c2_l1",
"landsat_tm_c2_l1"),
verbose=F) # NOTE: tm unavailable for toi_2018
# assign location for downloads
db.path <- "/media/seamus/Ubuntu 22_04 LTS amd64/rsat/edits/database"
ds.path <- "/media/seamus/Ubuntu 22_04 LTS amd64/rsat/edits/datasets"
dir.create(db.path)
dir.create(ds.path)Derive rtoi filter object for bugfree functions. This spatial fileter
object assigns download path for files and metadata, and allows to
restore and rapid plotting even after connection breaks. NOTE: better to
recall previous rtoi or can following chunk if previous rtoi was
purged.
# derive 'rtoi' object for spatial filter and handling results
chilwa_2018 <- new_rtoi(
name = "chilwa_2018",
region = ip,
db_path = db.path,
rtoi_path = ds.path)# apply temporal filter to rtoi collection
rsat_search(
region = chilwa_2018,
product = c("landsat_ot_c2_l1"),
dates = toi_2018)# download to rtoi-assigned folder
rsat_download(chilwa_2018)
# test_function(rsat_download)
# apply mosaic-clip-compression function
rsat_mosaic(chilwa_2018,
out_path = "/media/seamus/Ubuntu 22_04 LTS amd64/rsat/edits/mosaics/", # default to
warp = "extent",
region = ip,
overwrite = T)
# check results
rsat::plot(chilwa_2018,
as.Date("2018-01-11"),
band_name = c("nir", "red"))# convert to spatRaster
chilwa_2018_ndvi_rast = rsat_get_SpatRaster(
chilwa_2018, "landsat_ot_c2_l1", "NDVI")
# convert to RasterStack
chilwa_2018_ndvi_raster = rsat_get_raster(
chilwa_2018, "landsat_ot_c2_l1", "NDVI")
# convert to stars
chilwa_2018_ndvi_stars = rsat_get_stars(
chilwa_2018, "landsat_ot_c2_l1", "NDVI")
# extract single band in above formats
rsat_list_data(chilwa_2018)
# chilwa_2018_nir = rsat_get_SpatRaster(chilwa_2018,
# "landsat_ot_c2_l1",
# "LC81980182015183LGN00_B08")# derive custom function
NDVI = function(nir08, red){
ndvi <- (nir08 - red)/(nir08 + red)
return(ndvi)
}
# apply function to bands in rtoi collection
rsat_derive(chilwa_2018,
product = c("landsat_ot_c2_l1"),
variable = "ndvi", fun = NDVI)
# check results
plot(chilwa_2018,
as.Date("2018-11-11"),
variable = "ndvi",
xsize = 500,
ysize = 500,
zlim = c(-1,1))# alternatively, use pre-loaded spectral index
show_variables()
# replace with with variable of choice
rsat_derive(chilwa_2018, "NDWI", product = "landsat_ot_c2_l1")
rsat_list_data(chilwa_2018)
plot(chilwa_2018, as.Date("2018-11-11"),
variable = "ndvi",
xsize = 500,
ysize = 500,
zlim = c(-1,1))# register QA_Quality bands
rsat_cloudMask(chilwa_2018)
# derive NDSI layer and truncate to -1:1
ndsi.img <- rsat_get_raster(chilwa_2018, "landsat_ot_c2_l1", "ndsi")
ndsi.img <- clamp(ndsi.img, -1, 1)
# load function and mask results
clds.msk <- rsat_get_raster(chilwa_2018, "landsat_ot_c2_l1", "CloudMask")
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent