def download(asset, name, temporal_resolution, start_year, end_year,
EXECUTION_ID, logger):
"""
Download dataset from GEE assets.
"""
logger.debug("Entering download function.")
in_img = ee.Image(asset)
# if temporal_resolution != "one time":
# assert (start_year and end_year), "start year or end year not defined"
# out = in_img.select('y{}'.format(start_year))
# band_info = [BandInfo(name, add_to_map=True, metadata={'year': start_year})]
# for y in range(start_year + 1, end_year + 1):
# out.addBands(in_img.select('y{}'.format(start_year)))
# band_info.append(BandInfo(name, metadata={'year': start_year}))
# else:
# out = in_img
# band_info = [BandInfo(name, add_to_map=True)]
out = in_img
band_info = [BandInfo(name, add_to_map=True)]
n_bands = len(in_img.getInfo()['bands'])
if n_bands > 1:
band_info.extend([BandInfo(name, add_to_map=False)] * (n_bands - 1))
return TEImage(out, band_info)
def productivity_trajectory(year_start, year_end, method, ndvi_gee_dataset,
climate_gee_dataset, logger):
logger.debug("Entering productivity_trajectory function.")
climate_1yr = ee.Image(climate_gee_dataset)
climate_1yr = climate_1yr.where(climate_1yr.eq(9999), -32768)
climate_1yr = climate_1yr.updateMask(climate_1yr.neq(-32768))
if climate_gee_dataset == None and method != 'ndvi_trend':
raise GEEIOError("Must specify a climate dataset")
ndvi_dataset = ee.Image(ndvi_gee_dataset)
ndvi_dataset = ndvi_dataset.where(ndvi_dataset.eq(9999), -32768)
ndvi_dataset = ndvi_dataset.updateMask(ndvi_dataset.neq(-32768))
ndvi_mean = ndvi_dataset.select(ee.List(['y{}'.format(i) for i in range(year_start, year_end + 1)])) \
.reduce(ee.Reducer.mean()).rename(['ndvi'])
# Run the selected algorithm
if method == 'ndvi_trend':
lf_trend, mk_trend = ndvi_trend(year_start, year_end, ndvi_dataset, logger)
elif method == 'p_restrend':
lf_trend, mk_trend = p_restrend(year_start, year_end, ndvi_dataset, climate_1yr, logger)
if climate_1yr == None:
climate_1yr = precp_gpcc
elif method == 's_restrend':
#TODO: need to code this
raise GEEIOError("s_restrend method not yet supported")
elif method == 'ue':
lf_trend, mk_trend = ue_trend(year_start, year_end, ndvi_dataset, climate_1yr, logger)
else:
raise GEEIOError("Unrecognized method '{}'".format(method))
# Define Kendall parameter values for a significance of 0.05
period = year_end - year_start + 1
kendall90 = stats.get_kendall_coef(period, 90)
kendall95 = stats.get_kendall_coef(period, 95)
kendall99 = stats.get_kendall_coef(period, 99)
# Create final productivity trajectory output layer. Positive values are
# significant increase, negative values are significant decrease.
signif = ee.Image(-32768) \
.where(lf_trend.select('scale').gt(0).And(mk_trend.abs().gte(kendall90)), 1) \
.where(lf_trend.select('scale').gt(0).And(mk_trend.abs().gte(kendall95)), 2) \
.where(lf_trend.select('scale').gt(0).And(mk_trend.abs().gte(kendall99)), 3) \
.where(lf_trend.select('scale').lt(0).And(mk_trend.abs().gte(kendall90)), -1) \
.where(lf_trend.select('scale').lt(0).And(mk_trend.abs().gte(kendall95)), -2) \
.where(lf_trend.select('scale').lt(0).And(mk_trend.abs().gte(kendall99)), -3) \
.where(mk_trend.abs().lte(kendall90), 0) \
.where(lf_trend.select('scale').abs().lte(10), 0)
return TEImage(lf_trend.select('scale').addBands(signif).addBands(mk_trend).unmask(-32768).int16(),
[BandInfo("Productivity trajectory (trend)", metadata={'year_start': year_start, 'year_end': year_end}),
BandInfo("Productivity trajectory (significance)", add_to_map=True, metadata={'year_start': year_start, 'year_end': year_end}),
BandInfo("Mean annual NDVI integral", metadata={'year_start': year_start, 'year_end': year_end})])
def forest_gain(geojson, EXECUTION_ID, logger):
logger.debug("Entering Forest Gain function.")
gainImage = gfc2019.select(['gain'])
# version, for the clipping
poly = ee.Geometry(geojson, opt_geodesic=False)
gainImageAOI = gainImage.clip(poly)
gainAreaImage = gainImageAOI.multiply(ee.Image.pixelArea())
return TEImage(gainAreaImage.updateMask(gainAreaImage),
[BandInfo("Forest Gain", add_to_map=True, metadata={})])
def forest_loss(geojson, EXECUTION_ID, logger):
logger.debug("Entering Forest Loss Function.")
lossImage = gfc2019.select(['loss'])
# version, for the clipping
poly = ee.Geometry(geojson, opt_geodesic=False)
lossImageAOI = lossImage.clip(poly)
lossAreaImage = lossImageAOI.multiply(ee.Image.pixelArea())
return TEImage(lossAreaImage.updateMask(lossAreaImage),
[BandInfo("Forest Loss", add_to_map=True, metadata={})])
def forest_cover(year_start, year_end, ndvi_gee_dataset, geojson, EXECUTION_ID,
logger):
logger.debug("Entering Forest Cover function.")
treeCover = gfc2019.select(['treecover2000'])
# version, for the clipping
poly = ee.Geometry(geojson, opt_geodesic=False)
treeCoverAOI = treeCover.clip(poly)
treeCoverAreaImage = treeCoverAOI.multiply(ee.Image.pixelArea())
return TEImage(treeCoverAreaImage.updateMask(treeCoverAreaImage),
[BandInfo("Tree Cover", add_to_map=True, metadata={})])
def forest_loss_year(year, geojson, EXECUTION_ID, logger):
logger.debug("Entering Forest Loss Year function.")
# Make sure the bounding box of the poly is used, and not the geodesic
# version, for the clipping
poly = ee.Geometry(geojson, opt_geodesic=False)
lossYear = gfc2019.select(['lossyear']).eq(int(str(year)[2:]))
lossImageAOI = lossYear.clip(poly)
# calculate area in metres
lossAreaImage = lossImageAOI.multiply(ee.Image.pixelArea())
return TEImage(lossAreaImage.updateMask(lossAreaImage), [
BandInfo("Forest Loss in {0}".format(year),
add_to_map=True,
metadata={year: year})
])
def restoration_carbon(rest_type, length_yr, crs, geojsons, EXECUTION_ID,
logger):
logger.debug("Entering restoration_carbon function.")
# biomass
agb_30m = ee.Image(
"users/geflanddegradation/toolbox_datasets/forest_agb_30m_woodhole")
agb_1km = ee.Image(
"users/geflanddegradation/toolbox_datasets/forest_agb_1km_geocarbon")
# c sequestration coefficients by intervention from winrock paper
agfor0020 = ee.Image(
"users/geflanddegradation/toolbox_datasets/winrock_co2_seq_coeff_adm1"
).select("b1").divide(100)
agfor2060 = ee.Image(
"users/geflanddegradation/toolbox_datasets/winrock_co2_seq_coeff_adm1"
).select("b2").divide(100)
mshrr0020 = ee.Image(
"users/geflanddegradation/toolbox_datasets/winrock_co2_seq_coeff_adm1"
).select("b3").divide(100)
mshrr2060 = ee.Image(
"users/geflanddegradation/toolbox_datasets/winrock_co2_seq_coeff_adm1"
).select("b4").divide(100)
mtrer0020 = ee.Image(
"users/geflanddegradation/toolbox_datasets/winrock_co2_seq_coeff_adm1"
).select("b5").divide(100)
mtrer2060 = ee.Image(
"users/geflanddegradation/toolbox_datasets/winrock_co2_seq_coeff_adm1"
).select("b6").divide(100)
natre0020 = ee.Image(
"users/geflanddegradation/toolbox_datasets/winrock_co2_seq_coeff_adm1"
).select("b7").divide(100)
natre2060 = ee.Image(
"users/geflanddegradation/toolbox_datasets/winrock_co2_seq_coeff_adm1"
).select("b8").divide(100)
pwobr0020 = ee.Image(
"users/geflanddegradation/toolbox_datasets/winrock_co2_seq_coeff_adm1"
).select("b9").divide(100)
pweuc0020 = ee.Image(
"users/geflanddegradation/toolbox_datasets/winrock_co2_seq_coeff_adm1"
).select("b10").divide(100)
pwoak0020 = ee.Image(
"users/geflanddegradation/toolbox_datasets/winrock_co2_seq_coeff_adm1"
).select("b11").divide(100)
pwoco0020 = ee.Image(
"users/geflanddegradation/toolbox_datasets/winrock_co2_seq_coeff_adm1"
).select("b12").divide(100)
pwpin0020 = ee.Image(
"users/geflanddegradation/toolbox_datasets/winrock_co2_seq_coeff_adm1"
).select("b13").divide(100)
pwtea0020 = ee.Image(
"users/geflanddegradation/toolbox_datasets/winrock_co2_seq_coeff_adm1"
).select("b14").divide(100)
# combine the global 1km and the pantropical 30m datasets into one layer for analysis
agb = agb_30m.float().unmask(0).where(
ee.Image.pixelLonLat().select('latitude').abs().gte(29.999), agb_1km)
# calculate below ground biomass following Mokany et al. 2006 = (0.489)*(AGB)^(0.89)
bgb = agb.expression('0.489 * BIO**(0.89)', {'BIO': agb})
# calculate total biomass (t/ha) then convert to carbon equilavent (*0.5) to get Total Carbon (t ha-1) = (AGB+BGB)*0.5
tbc = agb.expression('(bgb + abg ) * 0.5 ', {'bgb': bgb, 'abg': agb})
# convert Total carbon to total CO2 eq (One ton of carbon equals 44/12 = 11/3 = 3.67 tons of carbon dioxide)
current_co2 = tbc.expression('totalcarbon * 3.67 ', {'totalcarbon': tbc})
if rest_type == "terrestrial":
if length_yr <= 20:
d_natre = natre0020.multiply(length_yr).subtract(current_co2)
d_natre = d_natre.where(d_natre.lte(0), 0)
d_agfor = agfor0020.multiply(length_yr).subtract(current_co2)
d_agfor = d_agfor.where(d_agfor.lte(0), 0)
d_pwtea = pwtea0020.multiply(length_yr).subtract(current_co2)
d_pweuc = pweuc0020.multiply(length_yr).subtract(current_co2)
d_pwoak = pwoak0020.multiply(length_yr).subtract(current_co2)
d_pwobr = pwobr0020.multiply(length_yr).subtract(current_co2)
d_pwpin = pwpin0020.multiply(length_yr).subtract(current_co2)
d_pwoco = pwoco0020.multiply(length_yr).subtract(current_co2)
if length_yr > 20:
d_natre = natre0020.multiply(20).add(
natre2060.multiply(length_yr - 20)).subtract(current_co2)
d_natre = d_natre.where(d_natre.lte(0), 0)
d_agfor = agfor0020.multiply(20).add(
agfor2060.multiply(length_yr - 20)).subtract(current_co2)
d_agfor = d_agfor.where(d_agfor.lte(0), 0)
d_pwtea = pwtea0020.multiply(20).subtract(current_co2)
d_pweuc = pweuc0020.multiply(20).subtract(current_co2)
d_pwoak = pwoak0020.multiply(20).subtract(current_co2)
d_pwobr = pwobr0020.multiply(20).subtract(current_co2)
d_pwpin = pwpin0020.multiply(20).subtract(current_co2)
d_pwoco = pwoco0020.multiply(20).subtract(current_co2)
output = current_co2.addBands(d_natre).addBands(d_agfor) \
.addBands(d_pwtea).addBands(d_pweuc) \
.addBands(d_pwoak).addBands(d_pwobr) \
.addBands(d_pwpin).addBands(d_pwoco) \
.rename(['current','natreg','agrfor','pwteak','pweuca','pwoaks','pwobro','pwpine','pwocon'])
logger.debug("Setting up output for terrestrial restoration.")
out = TEImage(output, [
BandInfo("Biomass (tonnes CO2e per ha)",
add_to_map=True,
metadata={'year': 'current'}),
BandInfo("Restoration biomass difference (tonnes CO2e per ha)",
metadata={
'years': length_yr,
'type': 'natural regeneration'
}),
BandInfo("Restoration biomass difference (tonnes CO2e per ha)",
metadata={
'years': length_yr,
'type': 'agroforestry'
}),
BandInfo("Restoration biomass difference (tonnes CO2e per ha)",
metadata={
'years': length_yr,
'type': 'teak plantation'
}),
BandInfo("Restoration biomass difference (tonnes CO2e per ha)",
metadata={
'years': length_yr,
'type': 'eucalyptus plantation'
}),
BandInfo("Restoration biomass difference (tonnes CO2e per ha)",
metadata={
'years': length_yr,
'type': 'oak plantation'
}),
BandInfo("Restoration biomass difference (tonnes CO2e per ha)",
metadata={
'years': length_yr,
'type': 'other broadleaf plantation'
}),
BandInfo("Restoration biomass difference (tonnes CO2e per ha)",
metadata={
'years': length_yr,
'type': 'pine plantation'
}),
BandInfo("Restoration biomass difference (tonnes CO2e per ha)",
metadata={
'years': length_yr,
'type': 'conifer plantation'
})
])
elif rest_type == "coastal":
if length_yr <= 20:
d_mshrr = mshrr0020.multiply(length_yr).subtract(current_co2)
d_mshrr = d_mshrr.where(d_mshrr.lte(0), 0)
d_mtrer = mtrer0020.multiply(length_yr).subtract(current_co2)
d_mtrer = d_mtrer.where(d_mtrer.lte(0), 0)
if length_yr > 20:
d_mshrr = mshrr0020.multiply(20).add(
mshrr2060.multiply(length_yr - 20)).subtract(current_co2)
d_mshrr = d_mshrr.where(d_mshrr.lte(0), 0)
d_mtrer = mtrer0020.multiply(20).add(
mtrer2060.multiply(length_yr - 20)).subtract(current_co2)
d_mtrer = d_mtrer.where(d_mtrer.lte(0), 0)
logger.debug("Setting up output for coastal restoration.")
out = TEImage(
current_co2.addBands(d_mshrr).addBands(d_mtrer).rename(
['current', 'mshrr', 'mtrer']),
[
BandInfo("Biomass (tonnes CO2e per ha)",
add_to_map=True,
metadata={'year': 'current'}),
BandInfo("Restoration biomass difference (tonnes CO2e per ha)",
metadata={
'years': length_yr,
'type': 'mangrove shrub'
}),
BandInfo("Restoration biomass difference (tonnes CO2e per ha)",
metadata={
'years': length_yr,
'type': 'mangrove tree'
})
])
else:
raise
out.image = out.image.reproject(crs=agb_30m.projection())
return out
def soc(geometry, year_start, year_end, fl, remap_matrix, dl_annual_lc,
EXECUTION_ID, logger):
"""
Calculate SOC indicator.
"""
logger.debug("Entering soc function.")
geom = ee.Geometry.Polygon(geometry)
# Location
area = ee.FeatureCollection(geom)
# soc
soc = ee.Image(
"users/geflanddegradation/toolbox_datasets/soc_sgrid_30cm").clip(area)
soc_t0 = soc.updateMask(soc.neq(-32768))
# land cover - note it needs to be reprojected to match soc so that it can
# be output to cloud storage in the same stack
lc = ee.Image("users/geflanddegradation/toolbox_datasets/lcov_esacc_1992_2018") \
.select(ee.List.sequence(year_start - 1992, year_end - 1992, 1)) \
.clip(area) \
.reproject(crs=soc.projection())
lc = lc.where(lc.eq(9999), -32768)
lc = lc.updateMask(lc.neq(-32768))
if fl == 'per pixel':
# Setup a raster of climate regimes to use for coding Fl automatically
climate = ee.Image("users/geflanddegradation/toolbox_datasets/ipcc_climate_zones") \
.clip(area) \
.remap([0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12],
[0, 2, 1, 2, 1, 2, 1, 2, 1, 5, 4, 4, 3])
clim_fl = climate.remap([0, 1, 2, 3, 4, 5],
[0, 0.8, 0.69, 0.58, 0.48, 0.64])
# create empty stacks to store annual land cover maps
stack_lc = ee.Image().select()
# create empty stacks to store annual soc maps
stack_soc = ee.Image().select()
# loop through all the years in the period of analysis to compute changes in SOC
for k in range(year_end - year_start):
# land cover map reclassified to UNCCD 7 classes (1: forest, 2:
# grassland, 3: cropland, 4: wetland, 5: artifitial, 6: bare, 7: water)
lc_t0 = lc.select(k).remap(remap_matrix[0], remap_matrix[1])
lc_t1 = lc.select(k + 1).remap(remap_matrix[0], remap_matrix[1])
if (k == 0):
# compute transition map (first digit for baseline land cover, and
# second digit for target year land cover)
lc_tr = lc_t0.multiply(10).add(lc_t1)
# compute raster to register years since transition
tr_time = ee.Image(2).where(lc_t0.neq(lc_t1), 1)
else:
# Update time since last transition. Add 1 if land cover remains
# constant, and reset to 1 if land cover changed.
tr_time = tr_time.where(lc_t0.eq(lc_t1), tr_time.add(ee.Image(1))) \
.where(lc_t0.neq(lc_t1), ee.Image(1))
# compute transition map (first digit for baseline land cover, and
# second digit for target year land cover), but only update where
# changes actually ocurred.
lc_tr_temp = lc_t0.multiply(10).add(lc_t1)
lc_tr = lc_tr.where(lc_t0.neq(lc_t1), lc_tr_temp)
# stock change factor for land use - note the 99 and -99 will be
# recoded using the chosen Fl option
lc_tr_fl_0 = lc_tr.remap([
11, 12, 13, 14, 15, 16, 17, 21, 22, 23, 24, 25, 26, 27, 31, 32, 33,
34, 35, 36, 37, 41, 42, 43, 44, 45, 46, 47, 51, 52, 53, 54, 55, 56,
57, 61, 62, 63, 64, 65, 66, 67, 71, 72, 73, 74, 75, 76, 77
], [
1, 1, 99, 1, 0.1, 0.1, 1, 1, 1, 99, 1, 0.1, 0.1, 1, -99, -99, 1,
1 / 0.71, 0.1, 0.1, 1, 1, 1, 0.71, 1, 0.1, 0.1, 1, 2, 2, 2, 2, 1,
1, 1, 2, 2, 2, 2, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1
])
if fl == 'per pixel':
lc_tr_fl = lc_tr_fl_0.where(lc_tr_fl_0.eq(99), clim_fl)\
.where(lc_tr_fl_0.eq(-99), ee.Image(1).divide(clim_fl))
else:
lc_tr_fl = lc_tr_fl_0.where(lc_tr_fl_0.eq(99), fl)\
.where(lc_tr_fl_0.eq(-99), ee.Image(1).divide(fl))
# stock change factor for management regime
lc_tr_fm = lc_tr.remap([
11, 12, 13, 14, 15, 16, 17, 21, 22, 23, 24, 25, 26, 27, 31, 32, 33,
34, 35, 36, 37, 41, 42, 43, 44, 45, 46, 47, 51, 52, 53, 54, 55, 56,
57, 61, 62, 63, 64, 65, 66, 67, 71, 72, 73, 74, 75, 76, 77
], [
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
1, 1, 1, 1, 1
])
# stock change factor for input of organic matter
lc_tr_fo = lc_tr.remap([
11, 12, 13, 14, 15, 16, 17, 21, 22, 23, 24, 25, 26, 27, 31, 32, 33,
34, 35, 36, 37, 41, 42, 43, 44, 45, 46, 47, 51, 52, 53, 54, 55, 56,
57, 61, 62, 63, 64, 65, 66, 67, 71, 72, 73, 74, 75, 76, 77
], [
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
1, 1, 1, 1, 1
])
if (k == 0):
soc_chg = (soc_t0.subtract((soc_t0.multiply(lc_tr_fl).multiply(
lc_tr_fm).multiply(lc_tr_fo)))).divide(20)
# compute final SOC stock for the period
soc_t1 = soc_t0.subtract(soc_chg)
# add to land cover and soc to stacks from both dates for the first
# period
stack_lc = stack_lc.addBands(lc_t0).addBands(lc_t1)
stack_soc = stack_soc.addBands(soc_t0).addBands(soc_t1)
else:
# compute annual change in soc (updates from previous period based
# on transition and time <20 years)
soc_chg = soc_chg.where(lc_t0.neq(lc_t1),
(stack_soc.select(k).subtract(stack_soc.select(k) \
.multiply(lc_tr_fl) \
.multiply(lc_tr_fm) \
.multiply(lc_tr_fo))).divide(20)) \
.where(tr_time.gt(20), 0)
# compute final SOC for the period
socn = stack_soc.select(k).subtract(soc_chg)
# add land cover and soc to stacks only for the last year in the
# period
stack_lc = stack_lc.addBands(lc_t1)
stack_soc = stack_soc.addBands(socn)
# compute soc percent change for the analysis period
soc_pch = ((stack_soc.select(year_end - year_start) \
.subtract(stack_soc.select(0))) \
.divide(stack_soc.select(0))) \
.multiply(100)
logger.debug("Setting up output.")
out = TEImage(soc_pch, [
BandInfo("Soil organic carbon (degradation)",
add_to_map=True,
metadata={
'year_start': year_start,
'year_end': year_end
})
])
logger.debug("Adding annual SOC layers.")
# Output all annual SOC layers
d_soc = []
for year in range(year_start, year_end + 1):
if (year == year_start) or (year == year_end):
add_to_map = True
else:
add_to_map = False
d_soc.append(
BandInfo("Soil organic carbon",
add_to_map=add_to_map,
metadata={'year': year}))
out.addBands(stack_soc, d_soc)
if dl_annual_lc:
logger.debug("Adding all annual LC layers.")
d_lc = []
for year in range(year_start, year_end + 1):
d_lc.append(
BandInfo("Land cover (7 class)", metadata={'year': year}))
out.addBands(stack_lc, d_lc)
else:
logger.debug("Adding initial and final LC layers.")
out.addBands(
stack_lc.select(0).addBands(
stack_lc.select(len(stack_lc.getInfo()['bands']) - 1)),
[
BandInfo("Land cover (7 class)", metadata={'year': year_start
}),
BandInfo("Land cover (7 class)", metadata={'year': year_end})
])
out.image = out.image.unmask(-32768).int16()
return out
def forest_fire(geometry,prefire_start,prefire_end,postfire_start,postfire_end, platform, EXECUTION_ID,logger):
"""
===========================================================================================
BURN SEVERITY MAPPING USING THE NORMALIZED BURN RATIO (NBR)
===========================================================================================
Normalized Burn Ratio will be applied to imagery from before and after a wild fire. By
calculating the difference afterwards (dNBR) Burn Severity is derived, showing the spatial
impact of the disturbance. Imagery used in this process comes from either Sentinel-2 or
Landsat 8.
"""
logger.debug("Entering forest_fire function.")
# SELECT one of the following: 'L8' or 'S2'
if platform == 'S2' or platform == 's2':
ImCol = 'COPERNICUS/S2'
pl = 'Sentinel-2'
else:
ImCol = 'LANDSAT/LC08/C01/T1_SR'
pl = 'Landsat 8'
# logger.debug(ee.String('Data selected for analysis: ').cat(pl))
# logger.debug(ee.String('Fire incident occurred between ').cat(prefire_end).cat(' and ').cat(postfire_start))
geom = ee.Geometry.Polygon(geometry)
# Location
area = ee.FeatureCollection(geom)
# add image collection
imagery = ee.ImageCollection(ImCol)
prefireImCol = ee.ImageCollection(imagery
# Filter by dates.
.filterDate(prefire_start, prefire_end)
# Filter by location.
.filterBounds(area))
postfireImCol = ee.ImageCollection(imagery
# Filter by dates.
.filterDate(postfire_start, postfire_end)
# Filter by location.
.filterBounds(area))
# logger.debug("Pre-fire Image Collection: "+prefireImCol)
# logger.debug("Post-fire Image Collection: "+postfireImCol)
def maskS2sr(image):
# Bits 10 and 11 are clouds and cirrus, respectively.
cloudBitMask = ee.Number(2).pow(10).int()
cirrusBitMask = ee.Number(2).pow(11).int()
# Get the pixel QA band.
qa = image.select('QA60')
# All flags should be set to zero, indicating clear conditions.
mask = qa.bitwiseAnd(cloudBitMask).eq(0).And(qa.bitwiseAnd(cirrusBitMask).eq(0))
# Return the masked image, scaled to TOA reflectance, without the QA bands.
return image.updateMask(mask).copyProperties(image, ["system:time_start"])
def maskL8sr(image):
# Bits 3 and 5 are cloud shadow and cloud, respectively.
cloudShadowBitMask = 1 << 3
cloudsBitMask = 1 << 5
snowBitMask = 1 << 4
# Get the pixel QA band.
qa = image.select('pixel_qa')
# All flags should be set to zero, indicating clear conditions.
mask = qa.bitwiseAnd(cloudShadowBitMask).eq(0).And(qa.bitwiseAnd(cloudsBitMask).eq(0)).And(qa.bitwiseAnd(snowBitMask).eq(0))
# Return the masked image, scaled to TOA reflectance, without the QA bands.
return image.updateMask(mask).select("B[0-9]*").copyProperties(image, ["system:time_start"])
# Apply platform-specific cloud mask
if platform == 'S2' or platform == 's2':
prefire_CM_ImCol = prefireImCol.map(maskS2sr)
postfire_CM_ImCol = postfireImCol.map(maskS2sr)
else:
prefire_CM_ImCol = prefireImCol.map(maskL8sr)
postfire_CM_ImCol = postfireImCol.map(maskL8sr)
pre_mos = prefireImCol.mosaic().clip(area)
post_mos = postfireImCol.mosaic().clip(area)
pre_cm_mos = prefire_CM_ImCol.mosaic().clip(area)
post_cm_mos = postfire_CM_ImCol.mosaic().clip(area)
if platform == 'S2' or platform == 's2':
preNBR = pre_cm_mos.normalizedDifference(['B8', 'B12'])
postNBR = post_cm_mos.normalizedDifference(['B8', 'B12'])
else:
preNBR = pre_cm_mos.normalizedDifference(['B5', 'B7'])
postNBR = post_cm_mos.normalizedDifference(['B5', 'B7'])
# The result is called delta NBR or dNBR
dNBR_unscaled = preNBR.subtract(postNBR)
# Scale product to USGS standards
dNBR = dNBR_unscaled.multiply(1000)
# reclassify dnbr
dNBR = dNBR \
.where(dNBR.gte(-500).And(dNBR.lte(-251)), 1) \
.where(dNBR.gte(-250).And(dNBR.lte(-101)), 2) \
.where(dNBR.gte(-100).And(dNBR.lte(99)), 3) \
.where(dNBR.gte(100).And(dNBR.lte(269)), 4) \
.where(dNBR.gte(270).And(dNBR.lte(439)), 5) \
.where(dNBR.gte(440).And(dNBR.lte(659)), 6) \
.where(dNBR.gte(660).And(dNBR.lte(1300)), 7) \
.rename("dNBR")
return TEImage(dNBR.addBands(preNBR).addBands(postNBR),
[BandInfo("dNBR image", add_to_map=True, metadata={'prefire_start':prefire_start,'prefire_end':prefire_end, 'postfire_start':postfire_start, 'postfire_end':postfire_end}),
BandInfo("Prefire Normalized Burn Ratio", add_to_map=True,metadata={'prefire_start':prefire_start,'prefire_end':prefire_end}),
BandInfo("Postfire Normalized Burn Ratio", add_to_map=True,metadata={'postfire_start':postfire_start, 'postfire_end':postfire_end})])
def urban(isi_thr, ntl_thr, wat_thr, cap_ope, pct_suburban, pct_urban, un_adju,
crs, geojsons, EXECUTION_ID, logger):
# Impervious surface index computed by Trends.Earth
isi_series = ee.ImageCollection("projects/trends_earth/isi_1998_2018_20190403").reduce(ee.Reducer.mean()) \
.select(['isi1998_mean','isi2000_mean','isi2005_mean','isi2010_mean','isi2015_mean','isi2018_mean'],
['isi1998','isi2000','isi2005','isi2010','isi2015','isi2018'])
proj = isi_series.select('isi2000').projection()
# JRC Global Surface Water Mapping Layers, v1.0 (>20% occurrence)
water = ee.Image("JRC/GSW1_0/GlobalSurfaceWater").select("occurrence")
# Define nighttime lights mask from VIIRS Nighttime Day/Night Band
# Composites Version 1 (Apr 1, 2012 - May 1, 2018)
ntl = ee.ImageCollection("NOAA/VIIRS/DNB/MONTHLY_V1/VCMCFG")
ntl = ntl.filterDate(ee.Date("2015-01-01"), ee.Date("2015-12-31")).select(["avg_rad"], ["ntl"]).median() \
.clip(ee.Geometry.Polygon([-180, 57, 0, 57, 180, 57, 180, -88, 0, -88, -180, -88], None, False)).unmask(10)
# Mask urban areas based ntl
urban98 = isi_series.select("isi1998").gte(isi_thr).unmask(0).where(ntl.lte(-1+ntl_thr*31/100), 0)
urban00 = isi_series.select("isi2000").gte(isi_thr).unmask(0).where(ntl.lte(-1+ntl_thr*31/100), 0)
urban05 = isi_series.select("isi2005").gte(isi_thr).unmask(0).where(ntl.lte(-1+ntl_thr*31/100), 0)
urban10 = isi_series.select("isi2010").gte(isi_thr).unmask(0).where(ntl.lte(-1+ntl_thr*31/100), 0)
urban15 = isi_series.select("isi2015").gte(isi_thr).unmask(0).where(ntl.lte(-1+ntl_thr*31/100), 0)
urban18 = isi_series.select("isi2018").gte(isi_thr).unmask(0).where(ntl.lte(-1+ntl_thr*31/100), 0)
urban_series = (urban98.multiply(100000)).add \
(urban00.multiply(10000)).add \
(urban05.multiply(1000)).add \
(urban10.multiply(100)).add \
(urban15.multiply(10)).add \
(urban18.multiply(1))
urban_sum = urban98.add(urban00).add(urban05).add(urban10).add(urban15).add(urban18)
if un_adju:
# Gridded Population of the World Version 4, UN-Adjusted Population
# Density
gpw4_2000 = ee.Image("CIESIN/GPWv4/unwpp-adjusted-population-density/2000") \
.select(["population-density"], ["p2000"]).reproject(crs=proj, scale=30)
gpw4_2005 = ee.Image("CIESIN/GPWv4/unwpp-adjusted-population-density/2005") \
.select(["population-density"], ["p2005"]).reproject(crs=proj, scale=30)
gpw4_2010 = ee.Image("CIESIN/GPWv4/unwpp-adjusted-population-density/2010") \
.select(["population-density"], ["p2010"]).reproject(crs=proj, scale=30)
gpw4_2015 = ee.Image("CIESIN/GPWv4/unwpp-adjusted-population-density/2015") \
.select(["population-density"], ["p2015"]).reproject(crs=proj, scale=30)
else:
gpw4_2000 = ee.Image("CIESIN/GPWv4/population-density/2000") \
.select(["population-density"], ["p2000"]).reproject(crs=proj, scale=30)
gpw4_2005 = ee.Image("CIESIN/GPWv4/population-density/2005") \
.select(["population-density"], ["p2005"]).reproject(crs=proj, scale=30)
gpw4_2010 = ee.Image("CIESIN/GPWv4/population-density/2010") \
.select(["population-density"], ["p2010"]).reproject(crs=proj, scale=30)
gpw4_2015 = ee.Image("CIESIN/GPWv4/population-density/2015") \
.select(["population-density"], ["p2015"]).reproject(crs=proj, scale=30)
urban_series = urban_series.where(urban_series.eq( 0), 0) \
.where(urban_series.eq( 1), 0) \
.where(urban_series.eq( 10), 1) \
.where(urban_series.eq( 11), 4) \
.where(urban_series.eq( 100), 0) \
.where(urban_series.eq( 101), 3) \
.where(urban_series.eq( 110), 3) \
.where(urban_series.eq( 111), 3) \
.where(urban_series.eq( 1000), 0) \
.where(urban_series.eq( 1001), 2) \
.where(urban_series.eq( 1010), 2) \
.where(urban_series.eq( 1011), 2) \
.where(urban_series.eq( 1100), 2) \
.where(urban_series.eq( 1101), 2) \
.where(urban_series.eq( 1110), 2) \
.where(urban_series.eq( 1111), 2) \
.where(urban_series.eq(10000), 0) \
.where(urban_series.eq(10001), 0) \
.where(urban_series.eq(10010), 0) \
.where(urban_series.eq(10011), 1) \
.where(urban_series.eq(10100), 0) \
.where(urban_series.eq(10101), 1) \
.where(urban_series.eq(10110), 1) \
.where(urban_series.eq(10111), 1) \
.where(urban_series.eq(11000), 0) \
.where(urban_series.eq(11001), 1) \
.where(urban_series.eq(11010), 1) \
.where(urban_series.eq(11011), 1) \
.where(urban_series.eq(11100), 1) \
.where(urban_series.eq(11101), 1) \
.where(urban_series.eq(11110), 1) \
.where(urban_series.eq(11111), 1) \
.where(urban_series.gte(100000).And(urban_sum.gte(3)), 1) \
.where(urban_series.gte(100000).And(urban_sum.lte(2)), 0) \
.where(water.gte(wat_thr), -1) \
.reproject(crs=proj, scale=30)
## define function to do zonation of cities
def f_city_zones(built_up, geojson):
dens = built_up.reduceNeighborhood(reducer=ee.Reducer.mean(), kernel=ee.Kernel.circle(564, "meters"))
##rural built up (-32768 no-data), suburban, urban
city = ee.Image(10).where(dens.lte(pct_suburban).And(built_up.eq(1)), 3) \
.where(dens.gt(pct_suburban).And(built_up.eq(1)), 2) \
.where(dens.gt(pct_urban).And(built_up.eq(1)), 1)
dist = city.lte(2).fastDistanceTransform(100).sqrt()
## fringe open space, rural built up
city = city.where(dist.gt(0).And(dist.lte(3)), 4) \
.where(city.eq(3), 3)
open_space = city.updateMask(city.eq(10)).addBands(ee.Image.pixelArea())
open_space_poly = open_space.reduceToVectors(
reducer=ee.Reducer.sum().setOutputs(['area']),
geometry=geojson,
scale=30,
maxPixels=1e10)
open_space_img = open_space_poly.reduceToImage(properties=['area'], reducer=ee.Reducer.first())
## captured open space, rural open space
city = city.where(city.eq(10).And(open_space_img.gt(0).And(open_space_img.lte(cap_ope*10000))), 5) \
.where(city.eq(10).And(open_space_img.gt(cap_ope*10000)), 6)
return city.where(city.eq(4).And(water.gte(wat_thr)), 7) \
.where(city.eq(5).And(water.gte(wat_thr)), 8) \
.where(city.eq(6).And(water.gte(wat_thr)), 9) \
.where(city.eq(10), -32768)
logger.debug("Processing geojsons")
outs = []
for geojson in geojsons:
city00 = f_city_zones(urban_series.eq(1), geojson)
city05 = f_city_zones(urban_series.gte(1).And(urban_series.lte(2)), geojson)
city10 = f_city_zones(urban_series.gte(1).And(urban_series.lte(3)), geojson)
city15 = f_city_zones(urban_series.gte(1).And(urban_series.lte(4)), geojson)
rast_export = city00.addBands(city05).addBands(city10).addBands(city15) \
.addBands(gpw4_2000).addBands(gpw4_2005).addBands(gpw4_2010).addBands(gpw4_2015).addBands(urban_series)
rast_export = rast_export.unmask(-32768).int16().reproject(crs=proj, scale=30)
this_out = TEImage(rast_export,
[BandInfo("Urban", metadata={'year': 2000}),
BandInfo("Urban", metadata={'year': 2005}),
BandInfo("Urban", metadata={'year': 2010}),
BandInfo("Urban", metadata={'year': 2015}),
BandInfo("Population", metadata={'year': 2000}),
BandInfo("Population", metadata={'year': 2005}),
BandInfo("Population", metadata={'year': 2010}),
BandInfo("Population", metadata={'year': 2015}),
BandInfo("Urban series", add_to_map=True)])
outs.append(this_out.export([geojson], 'urban', crs, logger, EXECUTION_ID))
return outs
def climate_quality(year, geometry, EXECUTION_ID, logger):
"""
===========================================================================================
CLIMATE QUALITY INDEX (CQI)
===========================================================================================
"""
logger.debug("Entering climate quality function.")
terra_climate = ee.ImageCollection("IDAHO_EPSCOR/TERRACLIMATE") \
.filter(ee.Filter.date('{}-01-01'.format(year), '{}-12-31'.format(year))) \
.mean() \
.clip(geometry)
precipitation = terra_climate.select('pr')
precipitation_class = precipitation \
.where(precipitation.gt(650), 1) \
.where(precipitation.gte(570).And(precipitation.lt(650)), 1.05) \
.where(precipitation.gte(490).And(precipitation.lt(570)), 1.15) \
.where(precipitation.gte(440).And(precipitation.lt(490)), 1.25) \
.where(precipitation.gte(390).And(precipitation.lt(440)), 1.35) \
.where(precipitation.gte(345).And(precipitation.lt(390)), 1.50) \
.where(precipitation.gte(310).And(precipitation.lt(345)), 1.65) \
.where(precipitation.gte(280).And(precipitation.lt(310)), 1.80) \
.where(precipitation.lt(280), 2) \
.rename("Precipitation")
evapotrans = terra_climate.select('pet')
aridityIndex = ee.Image().expression(
'(precipitation / evapotranspiration)', {
'precipitation': precipitation,
'evapotranspiration': evapotrans,
}).rename('aridityIndex')
aridityIndex = aridityIndex \
.where(aridityIndex.gte(1.0), 1) \
.where(aridityIndex.gte(0.75).And(aridityIndex.lt(1.0)), 1.05) \
.where(aridityIndex.gte(0.65).And(aridityIndex.lt(0.75)), 1.15) \
.where(aridityIndex.gte(0.50).And(aridityIndex.lt(0.65)), 1.25) \
.where(aridityIndex.gte(0.35).And(aridityIndex.lt(0.50)), 1.35) \
.where(aridityIndex.gte(0.20).And(aridityIndex.lt(0.35)), 1.45) \
.where(aridityIndex.gte(0.10).And(aridityIndex.lt(0.20)), 1.55) \
.where(aridityIndex.gte(0.03).And(aridityIndex.lt(0.10)), 1.75) \
.where(aridityIndex.lt(0.03), 2) \
.rename("Aridity Index")
cqi = ee.Image().expression('(precipitation * aridity_index) ** 1/2', {
'precipitation': precipitation_class,
'aridity_index': aridityIndex,
})
cqi_class = cqi \
.where(cqi.lt(1.15), 1) \
.where(cqi.gte(1.15).And(cqi.lte(1.81)), 2) \
.where(cqi.gt(1.81), 3) \
.rename('Climate Quality Reclass')
srtm_proj = ee.Image("USGS/SRTMGL1_003").projection()
# resample parent material to srtm projection
cqi_class = cqi_class.reproject(crs=srtm_proj)
return TEImage(cqi_class.clip(geometry), [
BandInfo("Climate Quality Index (year)",
add_to_map=True,
metadata={'year': year})
])
def land_cover(year_baseline, year_target, trans_matrix, remap_matrix,
EXECUTION_ID, logger):
"""
Calculate land cover indicator.
"""
logger.debug("Entering land_cover function.")
## land cover
lc = ee.Image(
"users/geflanddegradation/toolbox_datasets/lcov_esacc_1992_2018")
lc = lc.where(lc.eq(9999), -32768)
lc = lc.updateMask(lc.neq(-32768))
# Remap LC according to input matrix
lc_remapped = lc.select('y{}'.format(year_baseline)).remap(
remap_matrix[0], remap_matrix[1])
for year in range(year_baseline + 1, year_target + 1):
lc_remapped = lc_remapped.addBands(
lc.select('y{}'.format(year)).remap(remap_matrix[0],
remap_matrix[1]))
## target land cover map reclassified to IPCC 6 classes
lc_bl = lc_remapped.select(0)
## baseline land cover map reclassified to IPCC 6 classes
lc_tg = lc_remapped.select(len(lc_remapped.getInfo()['bands']) - 1)
## compute transition map (first digit for baseline land cover, and second digit for target year land cover)
lc_tr = lc_bl.multiply(10).add(lc_tg)
## definition of land cover transitions as degradation (-1), improvement (1), or no relevant change (0)
lc_dg = lc_tr.remap([
11, 12, 13, 14, 15, 16, 17, 21, 22, 23, 24, 25, 26, 27, 31, 32, 33, 34,
35, 36, 37, 41, 42, 43, 44, 45, 46, 47, 51, 52, 53, 54, 55, 56, 57, 61,
62, 63, 64, 65, 66, 67, 71, 72, 73, 74, 75, 76, 77
], trans_matrix)
## Remap persistence classes so they are sequential. This
## makes it easier to assign a clear color ramp in QGIS.
lc_tr = lc_tr.remap([
11, 12, 13, 14, 15, 16, 17, 21, 22, 23, 24, 25, 26, 27, 31, 32, 33, 34,
35, 36, 37, 41, 42, 43, 44, 45, 46, 47, 51, 52, 53, 54, 55, 56, 57, 61,
62, 63, 64, 65, 66, 67, 71, 72, 73, 74, 75, 76, 77
], [
1, 12, 13, 14, 15, 16, 17, 21, 2, 23, 24, 25, 26, 27, 31, 32, 3, 34,
35, 36, 37, 41, 42, 43, 4, 45, 46, 47, 51, 52, 53, 54, 5, 56, 57, 61,
62, 63, 64, 65, 6, 67, 71, 72, 73, 74, 75, 76, 7
])
logger.debug("Setting up output.")
out = TEImage(
lc_dg.addBands(lc.select('y{}'.format(year_baseline))).addBands(
lc.select('y{}'.format(year_target))).addBands(lc_tr), [
BandInfo("Land cover (degradation)",
add_to_map=True,
metadata={
'year_baseline': year_baseline,
'year_target': year_target
}),
BandInfo("Land cover (ESA classes)",
metadata={'year': year_baseline}),
BandInfo("Land cover (ESA classes)",
metadata={'year': year_target}),
BandInfo("Land cover transitions",
add_to_map=True,
metadata={
'year_baseline': year_baseline,
'year_target': year_target
})
])
# Return the full land cover timeseries so it is available for reporting
logger.debug("Adding annual lc layers.")
d_lc = []
for year in range(year_baseline, year_target + 1):
if (year == year_baseline) or (year == year_target):
add_to_map = True
else:
add_to_map = False
d_lc.append(
BandInfo("Land cover (7 class)",
add_to_map=add_to_map,
metadata={'year': year}))
out.addBands(lc_remapped, d_lc)
out.image = out.image.unmask(-32768).int16()
return out
def productivity_state(year_bl_start, year_bl_end,
year_tg_start, year_tg_end,
ndvi_gee_dataset, EXECUTION_ID, logger):
logger.debug("Entering productivity_state function.")
ndvi_1yr = ee.Image(ndvi_gee_dataset)
# compute min and max of annual ndvi for the baseline period
bl_ndvi_range = ndvi_1yr.select(ee.List(['y{}'.format(i) for i in range(year_bl_start, year_bl_end + 1)])) \
.reduce(ee.Reducer.percentile([0, 100]))
# add two bands to the time series: one 5% lower than min and one 5% higher than max
bl_ndvi_ext = ndvi_1yr.select(ee.List(['y{}'.format(i) for i in range(year_bl_start, year_bl_end + 1)])) \
.addBands(bl_ndvi_range.select('p0').subtract((bl_ndvi_range.select('p100').subtract(bl_ndvi_range.select('p0'))).multiply(0.05)))\
.addBands(bl_ndvi_range.select('p100').add((bl_ndvi_range.select('p100').subtract(bl_ndvi_range.select('p0'))).multiply(0.05)))
# compute percentiles of annual ndvi for the extended baseline period
percentiles = [10, 20, 30, 40, 50, 60, 70, 80, 90, 100]
bl_ndvi_perc = bl_ndvi_ext.reduce(ee.Reducer.percentile(percentiles))
# compute mean ndvi for the baseline and target period period
bl_ndvi_mean = ndvi_1yr.select(ee.List(['y{}'.format(i) for i in range(year_bl_start, year_bl_end + 1)])) \
.reduce(ee.Reducer.mean()).rename(['ndvi'])
tg_ndvi_mean = ndvi_1yr.select(ee.List(['y{}'.format(i) for i in range(year_tg_start, year_tg_end + 1)])) \
.reduce(ee.Reducer.mean()).rename(['ndvi'])
# reclassify mean ndvi for baseline period based on the percentiles
bl_classes = ee.Image(-32768) \
.where(bl_ndvi_mean.lte(bl_ndvi_perc.select('p10')), 1) \
.where(bl_ndvi_mean.gt(bl_ndvi_perc.select('p10')), 2) \
.where(bl_ndvi_mean.gt(bl_ndvi_perc.select('p20')), 3) \
.where(bl_ndvi_mean.gt(bl_ndvi_perc.select('p30')), 4) \
.where(bl_ndvi_mean.gt(bl_ndvi_perc.select('p40')), 5) \
.where(bl_ndvi_mean.gt(bl_ndvi_perc.select('p50')), 6) \
.where(bl_ndvi_mean.gt(bl_ndvi_perc.select('p60')), 7) \
.where(bl_ndvi_mean.gt(bl_ndvi_perc.select('p70')), 8) \
.where(bl_ndvi_mean.gt(bl_ndvi_perc.select('p80')), 9) \
.where(bl_ndvi_mean.gt(bl_ndvi_perc.select('p90')), 10)
# reclassify mean ndvi for target period based on the percentiles
tg_classes = ee.Image(-32768) \
.where(tg_ndvi_mean.lte(bl_ndvi_perc.select('p10')), 1) \
.where(tg_ndvi_mean.gt(bl_ndvi_perc.select('p10')), 2) \
.where(tg_ndvi_mean.gt(bl_ndvi_perc.select('p20')), 3) \
.where(tg_ndvi_mean.gt(bl_ndvi_perc.select('p30')), 4) \
.where(tg_ndvi_mean.gt(bl_ndvi_perc.select('p40')), 5) \
.where(tg_ndvi_mean.gt(bl_ndvi_perc.select('p50')), 6) \
.where(tg_ndvi_mean.gt(bl_ndvi_perc.select('p60')), 7) \
.where(tg_ndvi_mean.gt(bl_ndvi_perc.select('p70')), 8) \
.where(tg_ndvi_mean.gt(bl_ndvi_perc.select('p80')), 9) \
.where(tg_ndvi_mean.gt(bl_ndvi_perc.select('p90')), 10)
# difference between start and end clusters >= 2 means improvement (<= -2
# is degradation)
classes_chg = tg_classes.subtract(bl_classes).where(bl_ndvi_mean.subtract(tg_ndvi_mean).abs().lte(100), 0)
band_infos = [BandInfo("Productivity state (degradation)", add_to_map=True,
metadata={'year_bl_start': year_bl_start, 'year_bl_end': year_bl_end, 'year_tg_start': year_tg_start, 'year_tg_end': year_tg_end}),
BandInfo("Productivity state classes", metadata={'year_start': year_bl_start, 'year_end': year_bl_end}),
BandInfo("Productivity state classes", metadata={'year_start': year_tg_start, 'year_end': year_tg_end}),
BandInfo("Productivity state NDVI mean", metadata={'year_start': year_bl_start, 'year_end': year_bl_end}),
BandInfo("Productivity state NDVI mean", metadata={'year_start': year_tg_start, 'year_end': year_tg_end})]
return TEImage(classes_chg.addBands(bl_classes).addBands(tg_classes).addBands(bl_ndvi_mean).addBands(tg_ndvi_mean).int16(), band_infos)
def management_quality(year, lu_matrix, geometry, EXECUTION_ID, logger):
logger.debug("Entering management quality function.")
lu_remap_matrix = [[
10, 11, 12, 20, 30, 40, 50, 60, 61, 62, 70, 71, 72, 80, 81, 82, 90,
100, 110, 120, 121, 122, 130, 140, 150, 151, 152, 153, 160, 170, 180,
200, 201, 202
], lu_matrix]
## land cover
lc = ee.Image(
"users/geflanddegradation/toolbox_datasets/lcov_esacc_1992_2018").clip(
geometry)
lc = lc.where(lc.eq(9999), -32768)
lc = lc.updateMask(lc.neq(-32768))
# DEFINE LAND USE INTENSITY
# Remap LC according to input matrix
lc_remapped_lu = lc.select('y' + '{}'.format(year)).remap(
lu_remap_matrix[0], lu_remap_matrix[1]).divide(10)
# DEFINE POPULATION INTENSITY
population_density = ee.ImageCollection(
"CIESIN/GPWv411/GPW_Population_Density").toList(5)
if (year <= 2000):
population_density = ee.Image(population_density.get(0))
elif (year > 2000 and year <= 2005):
population_density = ee.Image(population_density.get(1))
elif (year > 2005 and year <= 2010):
population_density = ee.Image(population_density.get(2))
elif (year > 2010 and year <= 2015):
population_density = ee.Image(population_density.get(3))
elif (year > 2015):
population_density = ee.Image(population_density.get(4))
else:
logger.debug("Invalid date range")
population_density_class = population_density \
.select("population_density") \
.where(population_density.lt(4), 1.0) \
.where(population_density.gte(4).And(population_density.lt(30)), 1.1) \
.where(population_density.gte(30).And(population_density.lt(80)), 1.2) \
.where(population_density.gte(80).And(population_density.lt(170)), 1.3) \
.where(population_density.gte(170).And(population_density.lt(300)), 1.4) \
.where(population_density.gte(300).And(population_density.lt(500)), 1.5) \
.where(population_density.gte(500).And(population_density.lt(850)), 1.6) \
.where(population_density.gte(850).And(population_density.lt(1400)), 1.7) \
.where(population_density.gte(1400).And(population_density.lt(2000)), 1.8) \
.where(population_density.gte(2000).And(population_density.lt(2700)), 1.9) \
.where(population_density.gte(2700), 2.0) \
.clip(geometry)
mqi = ee.Image()
mqi = mqi.expression(
'(landuse_intensity * population_density)**(1/2)', {
'landuse_intensity': lc_remapped_lu,
'population_density': population_density_class
}).rename("Management Quality Index")
mqi_range = mqi \
.where(mqi.lte(1.25), 1) \
.where(mqi.lte(1.50).And(mqi.gt(1.25)), 2) \
.where(mqi.gt(1.50), 3)
return TEImage(mqi_range.clip(geometry), [
BandInfo("Management Quality Index",
add_to_map=True,
metadata={'year': year})
])
def tc(geometry, fc_threshold, year_start, year_end, method, biomass_data,
EXECUTION_ID, logger):
"""
Calculate total carbon (in belowground and aboveground biomass).
"""
logger.debug("Entering tc function.")
# geom = ee.Geometry.Polygon(geometry)
# Location
# area = ee.FeatureCollection(geom)
##############################################
# DATASETS
# Import Hansen global forest dataset
hansen = ee.Image('UMD/hansen/global_forest_change_2019_v1_7').clip(
geometry)
# Aboveground Live Woody Biomass per Hectare (Mg/Ha)
if biomass_data == 'woodshole':
agb = ee.ImageCollection(
"users/geflanddegradation/toolbox_datasets/forest_agb_30m_gfw"
).mosaic().unmask(0)
elif biomass_data == 'geocarbon':
agb = ee.Image(
"users/geflanddegradation/toolbox_datasets/forest_agb_1km_geocarbon"
).clip(geometry)
else:
agb = None
# All datasets will be reprojected to Hansen resolution
agb = agb.reproject(crs=hansen.projection())
# JRC Global Surface Water Mapping Layers, v1.0 (>50% occurrence)
water = ee.Image("JRC/GSW1_0/GlobalSurfaceWater").select(
"occurrence").clip(geometry)
water = water.reproject(crs=hansen.projection())
# reclass to 1.broadleaf, 2.conifer, 3.mixed, 4.savanna
f_type = ee.Image("users/geflanddegradation/toolbox_datasets/esa_forest_expanded_2015") \
.clip(geometry) \
.remap([50,60,61,62,70,71,72,80,81,82,90,100,110],
[ 1, 1, 1, 1, 2, 2, 2, 2, 2, 2, 3, 3, 3])
f_type = f_type.reproject(crs=hansen.projection())
# IPCC climate zones reclassified as from http://eusoils.jrc.ec.europa.eu/projects/RenewableEnergy/
# 0-No data, 1-Warm Temperate Moist, 2-k Temperate Dry, 3-Cool Temperate
# Moist, 4-Cool Temperate Dry, 5-Polar Moist,
# 6-Polar Dry, 7-Boreal Moist, 8-Boreal Dry, 9-Tropical Montane, 10-Tropical Wet, 11-Tropical Moist, 12-Tropical Dry) to
# 0: no data, 1:trop/sub moist, 2: trop/sub dry, 3: temperate)
climate = ee.Image("users/geflanddegradation/toolbox_datasets/ipcc_climate_zones") \
.clip(geometry) \
.remap([0,1,2,3,4,5,6,7,8,9,10,11,12],
[0,1,2,3,3,3,3,3,3,1, 1, 1, 2])
climate = climate.reproject(crs=hansen.projection())
# Root to shoot ratio methods
if method == 'ipcc':
rs_ratio = (
ee.Image(-32768)
# low biomass wet tropical forest
.where(climate.eq(1).And(agb.lte(125)), 0.42)
# high biomass wet tropical forest
.where(climate.eq(1).And(agb.gte(125)), 0.24)
# dry tropical forest
.where(climate.eq(2), 0.27)
# low biomass temperate conifer forest
.where(climate.eq(3).And(f_type.eq(2).And(agb.lte(50))), 0.46)
# mid biomass temperate conifer forest
.where(
climate.eq(3).And(
f_type.eq(2).And(agb.gte(50)).And(agb.lte(150))), 0.32)
# high biomass temperate conifer forest
.where(climate.eq(3).And(f_type.eq(2).And(agb.lte(150))), 0.23)
# low biomass temperate broadleaf forest
.where(climate.eq(3).And(f_type.eq(1).And(agb.lte(75))), 0.43)
# low biomass temperate broadleaf forest
.where(
climate.eq(3).And(
f_type.eq(1).And(agb.gte(75)).And(agb.lte(150))), 0.26)
# low biomass temperate broadleaf forest
.where(climate.eq(3).And(f_type.eq(1).And(agb.lte(150))), 0.24)
# low biomass temperate mixed forest
.where(
climate.eq(3).And(f_type.eq(1).And(agb.lte(75))),
(0.46 + 0.43) / 2)
# low biomass temperate mixed forest
.where(
climate.eq(3).And(
f_type.eq(1).And(agb.gte(75)).And(agb.lte(150))),
(0.32 + 0.26) / 2)
# low biomass temperate mixed forest
.where(
climate.eq(3).And(f_type.eq(1).And(agb.lte(150))),
(0.23 + 0.24) / 2)
# savanas regardless of climate
.where(f_type.eq(4), 2.8))
bgb = agb.multiply(rs_ratio)
elif (method == 'mokany'):
# calculate average above and below ground biomass
# BGB (t ha-1) Citation Mokany et al. 2006 = (0.489)*(AGB)^(0.89)
# Mokany used a linear regression of root biomass to shoot biomass for forest
# and woodland and found that BGB(y) is ~ 0.489 of AGB(x). However,
# applying a power (0.89) to the shoot data resulted in an improved
# model for relating root biomass (y) to shoot biomass (x):
# y = 0:489 x0:890
bgb = agb.expression('0.489 * BIO**(0.89)', {'BIO': agb})
rs_ratio = bgb.divide(agb)
else:
raise
# Calculate Total biomass (t/ha) then convert to carbon equilavent (*0.5) to get Total Carbon (t ha-1) = (AGB+BGB)*0.5
tbcarbon = agb.expression('(bgb + abg) * 0.5', {'bgb': bgb, 'abg': agb})
# convert Total Carbon to Total Carbon dioxide tCO2/ha
# One ton of carbon equals 44/12 = 11/3 = 3.67 tons of carbon dioxide
# teco2 = agb.expression('totalcarbon * 3.67 ', {'totalcarbon': tbcarbon})
##############################################/
# define forest cover at the starting date
fc_str = hansen.select("treecover2000").gte(fc_threshold) \
.multiply(hansen.select('lossyear').unmask(0).lte(0).add(hansen.select('lossyear').unmask(0).gt(year_start-2000)))
# Create three band layer clipped to study area
# Band 1: forest layer for initial year (0) and year loss coded as numbers (e.g. 1 = 2001)
# Band 2: root to shoot ratio
# Band 3: total carbon stocks (tons of C per ha)
output = fc_str.multiply(hansen.select('lossyear').unmask(0) \
.lte(year_end-2000).multiply(hansen.select('lossyear').unmask(0))) \
.where(fc_str.eq(0),-1) \
.where(water.gte(50),-2).unmask(-32768) \
.addBands((rs_ratio.multiply(100)).multiply(fc_str)).unmask(-32768) \
.addBands((tbcarbon.multiply(10)).multiply(fc_str)).unmask(-32768)
output = output.reproject(crs=hansen.projection())
logger.debug("Setting up output.")
out = TEImage(output.int16().clip(geometry), [
BandInfo("Forest loss",
add_to_map=True,
metadata={
'year_start': year_start,
'year_end': year_end,
'ramp_min': year_start - 2000 + 1,
'ramp_max': year_end - 2000,
'threshold': fc_threshold
}),
BandInfo(
"Root/shoot ratio", add_to_map=False, metadata={'method': method}),
BandInfo("Total carbon",
add_to_map=True,
metadata={
'year_start': year_start,
'year_end': year_end,
'method': method,
'threshold': fc_threshold
})
])
return out
def urban_area(geojson, un_adju, EXECUTION_ID, logger):
"""
Calculate urban area.
"""
logger.debug("Entering urban_area function.")
aoi = ee.Geometry(geojson)
# Read asset with the time series of urban extent
urban_series = ee.Image("users/geflanddegradation/toolbox_datasets/urban_series").int32()
# Load population data from GPWv4: Gridded Population of the World Version 4, People/km2 (not UN adjusted)
if un_adju:
pop_densi = ee.ImageCollection("CIESIN/GPWv4/population-density")
else:
pop_densi = ee.ImageCollection("CIESIN/GPWv4/unwpp-adjusted-population-density")
# the .multiply(100).int32() is to make the rasters integers, with units for density of people/km2 * 100
# Select population datasets for each year
pop2000 = pop_densi.filter(ee.Filter.eq("system:index", "2000")).mean().multiply(100).int32()
pop2005 = pop_densi.filter(ee.Filter.eq("system:index", "2005")).mean().multiply(100).int32()
pop2010 = pop_densi.filter(ee.Filter.eq("system:index", "2010")).mean().multiply(100).int32()
pop2015 = pop_densi.filter(ee.Filter.eq("system:index", "2015")).mean().multiply(100).int32()
# Use urban extent to mask population density data
urb_pop2000 = pop2000.updateMask(urban_series.eq(1))
urb_pop2005 = pop2005.updateMask(urban_series.gte(1).And(urban_series.lte(2)))
urb_pop2010 = pop2010.updateMask(urban_series.gte(1).And(urban_series.lte(3)))
urb_pop2015 = pop2015.updateMask(urban_series.gte(1).And(urban_series.lte(4)))
# Compute mean population density per year
urb_pop2000m = urb_pop2000.reduceRegion(reducer=ee.Reducer.mean(),
geometry=aoi, scale=30,
maxPixels=1e12)
urb_pop2005m = urb_pop2005.reduceRegion(reducer= ee.Reducer.mean(),
geometry=aoi, scale=30,
maxPixels=1e12)
urb_pop2010m = urb_pop2010.reduceRegion(reducer=ee.Reducer.mean(),
geometry=aoi, scale=30,
maxPixels=1e12)
urb_pop2015m = urb_pop2015.reduceRegion(reducer=ee.Reducer.mean(),
geometry=aoi, scale=30,
maxPixels=1e12)
# Compute urban area per year
pixel_area = urban_series.updateMask(urban_series.eq(1)).multiply(ee.Image.pixelArea())
urb_are2000 = urban_series.updateMask(urban_series.eq(1)).multiply(ee.Image.pixelArea()) \
.reduceRegion(reducer=ee.Reducer.sum(), geometry=aoi, scale=30,
maxPixels=1e12)
urb_are2005 = urban_series.updateMask(urban_series.gte(1).And(urban_series.lte(2))).multiply(ee.Image.pixelArea()) \
.reduceRegion(reducer=ee.Reducer.sum(), geometry=aoi, scale=30,
maxPixels=1e12)
urb_are2010 = urban_series.updateMask(urban_series.gte(1).And(urban_series.lte(3))).multiply(ee.Image.pixelArea()) \
.reduceRegion(reducer=ee.Reducer.sum(), geometry=aoi, scale=30,
maxPixels=1e12)
urb_are2015 = urban_series.updateMask(urban_series.gte(1).And(urban_series.lte(4))).multiply(ee.Image.pixelArea()) \
.reduceRegion(reducer=ee.Reducer.sum(), geometry=aoi, scale=30,
maxPixels=1e12)
# Make a dictionary to contain results
result_table = ee.Dictionary({
"pop_dens2000": urb_pop2000m.get('population-density'),
"pop_dens2005": urb_pop2005m.get('population-density'),
"pop_dens2010": urb_pop2010m.get('population-density'),
"pop_dens2015": urb_pop2015m.get('population-density'),
"urb_area2000": urb_are2000.get('classification'),
"urb_area2005": urb_are2005.get('classification'),
"urb_area2010": urb_are2010.get('classification'),
"urb_area2015": urb_are2015.get('classification'),
})
# # Export the FeatureCollection.
# Export.table.toDrive({
# collection: ee.FeatureCollection([ee.Feature(null,result_table)]),
# description: "export_urban_extent_table",
# folder: 'sdg1131',
# fileFormat: 'CSV'})
#
# Export raster
result_raster = urban_series.addBands(urb_pop2000).addBands(urb_pop2005).addBands(urb_pop2010).addBands(urb_pop2015)
logger.debug("Setting up output.")
out = TEImage(result_raster.clip(aoi),
[BandInfo("Urban series", add_to_map=True, metadata={'years': [2000, 2005, 2010, 2015]}),
BandInfo("Population", metadata={'year': 2000}),
BandInfo("Population", metadata={'year': 2005}),
BandInfo("Population", metadata={'year': 2010}),
BandInfo("Population", add_to_map=True, metadata={'year': 2015})])
#out.image = out.image.unmask(-32768).int16()
return out
def soil_quality(depth, texture_matrix, geometry, EXECUTION_ID, logger):
"""
===========================================================================================
SOIL QUALITY INDEX (SQI)
===========================================================================================
The impact of the soil factor to the process of desertification is determined by the strength of cohesion between soil particles, water retention ability of the soil, soil texture, and structure. The soil quality index (SQI), developed by OSS, that will be used is based on four parameters:
- parent material
- soil depth (0 to 200cm)
- soil texture
- slope
- rock fragment
- drainage
The formula used to compute the SQI from the above-mentioned parameters is as shown below:
SQI = (parent material*soil depth*soil texture* slope* rock fragment*drainage)^1/6
"""
logger.debug("Entering soil quality function.")
srtm = ee.Image("USGS/SRTMGL1_003")
# ==========================
# PARENT MATERIAL
# ==========================
# parent_material_map = [
# [1,2,3,4,5,6,7,8,9, 10, 11, 12, 13, 14, 15],pmaterial_matrix]
parent_material = ee.Image("users/miswagrace/parent_material_northafrica").clip(geometry)
parent_material = parent_material.remap([1.0, 1.2, 1.4, 1.5, 1.6, 1.7, 2.0],[1.0, 1.7, 1.7, 1.7, 1.7, 1.7, 2.0])
# resample parent material to srtm projection
parent_material = parent_material.reproject(crs=srtm.projection())
# remap parent material
# parent_material = parent_material.remap(parent_material_map[0], parent_material_map[1])
# ==========================
# SLOPE
# ==========================
slope = ee.Image("users/miswagrace/slope_north_africa").clip(geometry)
slope = slope.reproject(crs=srtm.projection())
# ==========================
# TEXTURE
# ==========================
# soil_texture = ee.Image("OpenLandMap/SOL/SOL_TEXTURE-CLASS_USDA-TT_M/v02").clip(geometry)
soil_texture = ee.Image("users/miswagrace/texture_north_africa").clip(geometry)
remap_texture = [
[1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13],texture_matrix]
# remap texture
soil_texture_remap = soil_texture.remap(remap_texture[0], remap_texture[1])
# soil_texture_remap = soil_texture_remap \
# .where(soil_texture_remap.eq(1), 1) \
# .where(soil_texture_remap.eq(2), 1.2) \
# .where(soil_texture_remap.eq(3), 1.6) \
# .where(soil_texture_remap.eq(4), 2) \
# .rename("Soil Texture")
# soil_texture_remap = soil_texture_remap.updateMask(soil_texture_remap.neq(-32768))
# ==========================
# DEPTH
# ==========================
if depth<15 :
depthIndex = 4
elif depth>=15 and depth<30:
depthIndex = 3
elif depth>=30 and depth<75:
depthIndex = 2
elif depth>=75:
depthIndex = 1
else:
logger.debug("Unexpected depth value")
# ==========================
# ROCK FRAGMENT
# ==========================
# classify rock fragment based on 3 classes
rock_fragment = ee.Image("users/miswagrace/rock_fragment_north_africa")
fragmentClass = rock_fragment \
.where(rock_fragment.lt(20), 2) \
.where(rock_fragment.gte(20).And(rock_fragment.lte(60)), 1.3) \
.where(rock_fragment.gt(60), 1) \
.rename("Rock Fragments")
# ==========================
# SOIL DRAINAGE
# ==========================
drainage = ee.Image("users/miswagrace/drainage_northafrica")
soil_drainage = drainage.remap([1,2,3,4,5,6,7],[2.0, 2.0, 1.4, 1.2, 1.0, 1.7, 2.0])
# ==========================
# SOIL QUALITY INDEX
# ==========================
# compute soil quality index based on datasetes generated above
img = ee.Image()
sqi = img.expression('(slope * parent_material * soil_texture * soil_depth * rock_fragment *drainage) ** (1/6)', {
'slope':slope,
'parent_material':parent_material,
'soil_texture':soil_texture_remap,
'soil_depth':depthIndex,
'rock_fragment':fragmentClass.clip(geometry),
'drainage':soil_drainage.clip(geometry)
}).rename('sqi')
# classify output sqi into 3 classes
sqi = sqi \
.where(sqi.lt(1.13), 1) \
.where(sqi.gte(1.13).And(sqi.lte(1.45)), 2) \
.where(sqi.gt(1.45), 3)
return TEImage(sqi,
[BandInfo("Soil Quality Index (cm deep)", add_to_map=True, metadata={'depth':depth})])
def vegetation_quality(year, ndvi_start, ndvi_end, drought_matrix, fire_matrix,
erosion_matrix, geometry, EXECUTION_ID, logger):
drought_remap_matrix = [[
10, 11, 12, 20, 30, 40, 50, 60, 61, 62, 70, 71, 72, 80, 81, 82, 90,
100, 110, 120, 121, 122, 130, 140, 150, 151, 152, 153, 160, 170, 180,
200, 201, 202
], drought_matrix]
fire_remap_matrix = [[
10, 11, 12, 20, 30, 40, 50, 60, 61, 62, 70, 71, 72, 80, 81, 82, 90,
100, 110, 120, 121, 122, 130, 140, 150, 151, 152, 153, 160, 170, 180,
200, 201, 202
], fire_matrix]
erosion_remap_matrix = [[
10, 11, 12, 20, 30, 40, 50, 60, 61, 62, 70, 71, 72, 80, 81, 82, 90,
100, 110, 120, 121, 122, 130, 140, 150, 151, 152, 153, 160, 170, 180,
200, 201, 202
], erosion_matrix]
## land cover
lc = ee.Image(
"users/geflanddegradation/toolbox_datasets/lcov_esacc_1992_2018").clip(
geometry)
lc = lc.where(lc.eq(9999), -32768)
lc = lc.updateMask(lc.neq(-32768))
# DEFINE DROUGHT RESISTANCE
lc_remapped_drought = lc.select('y' + year).remap(
drought_remap_matrix[0], drought_remap_matrix[1]).divide(10)
# DEFINE FIRE RISK
lc_remapped_fire = lc.select('y' + year).remap(
fire_remap_matrix[0], fire_remap_matrix[1]).divide(10)
# DEFINE EROSION PROTECTION
lc_remapped_erosion = lc.select('y' + year).remap(
erosion_remap_matrix[0], erosion_remap_matrix[1]).divide(10)
# DEFINE PLANT COVER
max_ndvi = ee.ImageCollection('VITO/PROBAV/C1/S1_TOC_100M') \
.filter(ee.Filter.date(ndvi_start, ndvi_end)) \
.max() \
.select('NDVI')
plant_cover_range = max_ndvi.divide(255).multiply(100)
plant_cover_class = plant_cover_range \
.where(plant_cover_range.gte(80), 1.0) \
.where(plant_cover_range.lt(80).And(plant_cover_range.gte(72)), 1.1) \
.where(plant_cover_range.lt(72).And(plant_cover_range.gte(62)), 1.2) \
.where(plant_cover_range.lt(62).And(plant_cover_range.gte(50)), 1.3) \
.where(plant_cover_range.lt(50).And(plant_cover_range.gte(38)), 1.4) \
.where(plant_cover_range.lt(38).And(plant_cover_range.gte(16)), 1.5) \
.where(plant_cover_range.lt(26).And(plant_cover_range.gte(18)), 1.6) \
.where(plant_cover_range.lt(18).And(plant_cover_range.gte(13)), 1.7) \
.where(plant_cover_range.lt(13).And(plant_cover_range.gte(11)), 1.8) \
.where(plant_cover_range.lt(11).And(plant_cover_range.gte(10)), 1.9) \
.where(plant_cover_range.lt(10), 2) \
.clip(geometry) \
.rename("Plant Cover")
# CALCULATE VEGETATION QUALITY INDEX
vqi = ee.Image()
vqi = vqi.expression(
'(fire_risk * erosion_protection * drought_resistance * plant_cover)**(1/4)',
{
'fire_risk': lc_remapped_fire,
'erosion_protection': lc_remapped_erosion,
'drought_resistance': lc_remapped_drought,
'plant_cover': plant_cover_class,
}).rename("Vegetation Quality Index")
vqi_range = vqi \
.where(vqi.lte(1.13), 1) \
.where(vqi.lte(1.38).And(vqi.gt(1.13)), 2) \
.where(vqi.gt(1.38), 3)
return TEImage(vqi_range.clip(geometry), [
BandInfo("Vegetation Quality Index",
add_to_map=True,
metadata={'year': year})
])
def productivity_performance(year_start, year_end, ndvi_gee_dataset, geojson,
EXECUTION_ID, logger):
logger.debug("Entering productivity_performance function.")
ndvi_1yr = ee.Image(ndvi_gee_dataset)
ndvi_1yr = ndvi_1yr.where(ndvi_1yr.eq(9999), -32768)
ndvi_1yr = ndvi_1yr.updateMask(ndvi_1yr.neq(-32768))
# land cover data from esa cci
lc = ee.Image("users/geflanddegradation/toolbox_datasets/lcov_esacc_1992_2019")
lc = lc.where(lc.eq(9999), -32768)
lc = lc.updateMask(lc.neq(-32768))
# global agroecological zones from IIASA
soil_tax_usda = ee.Image("users/geflanddegradation/toolbox_datasets/soil_tax_usda_sgrid")
# Make sure the bounding box of the poly is used, and not the geodesic
# version, for the clipping
poly = ee.Geometry(geojson, opt_geodesic=False)
# compute mean ndvi for the period
ndvi_avg = ndvi_1yr.select(ee.List(['y{}'.format(i) for i in range(year_start, year_end + 1)])) \
.reduce(ee.Reducer.mean()).rename(['ndvi']).clip(poly)
# Handle case of year_start that isn't included in the CCI data
if year_start > 2015:
lc_year_start = 2015
elif year_start < 1992:
lc_year_start = 1992
else:
lc_year_start = year_start
# reclassify lc to ipcc classes
lc_t0 = lc.select('y{}'.format(lc_year_start)) \
.remap([10, 11, 12, 20, 30, 40, 50, 60, 61, 62, 70, 71, 72, 80, 81, 82, 90, 100, 160, 170, 110, 130, 180, 190, 120, 121, 122, 140, 150, 151, 152, 153, 200, 201, 202, 210],
[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])
# create a binary mask.
mask = ndvi_avg.neq(0)
# define modis projection attributes
modis_proj = ee.Image("users/geflanddegradation/toolbox_datasets/ndvi_modis_2001_2020").projection()
# reproject land cover, soil_tax_usda and avhrr to modis resolution
lc_proj = lc_t0.reproject(crs=modis_proj)
soil_tax_usda_proj = soil_tax_usda.reproject(crs=modis_proj)
ndvi_avg_proj = ndvi_avg.reproject(crs=modis_proj)
# define unit of analysis as the intersect of soil_tax_usda and land cover
units = soil_tax_usda_proj.multiply(100).add(lc_proj)
# create a 2 band raster to compute 90th percentile per unit (analysis restricted by mask and study area)
ndvi_id = ndvi_avg_proj.addBands(units).updateMask(mask)
# compute 90th percentile by unit
perc90 = ndvi_id.reduceRegion(reducer=ee.Reducer.percentile([90]).
group(groupField=1, groupName='code'),
geometry=poly,
scale=ee.Number(modis_proj.nominalScale()).getInfo(),
maxPixels=1e15)
# Extract the cluster IDs and the 90th percentile
groups = ee.List(perc90.get("groups"))
ids = groups.map(lambda d: ee.Dictionary(d).get('code'))
perc = groups.map(lambda d: ee.Dictionary(d).get('p90'))
# remap the units raster using their 90th percentile value
raster_perc = units.remap(ids, perc)
# compute the ration of observed ndvi to 90th for that class
obs_ratio = ndvi_avg_proj.divide(raster_perc)
# aggregate obs_ratio to original NDVI data resolution (for modis this step does not change anything)
obs_ratio_2 = obs_ratio.reduceResolution(reducer=ee.Reducer.mean(), maxPixels=2000) \
.reproject(crs=ndvi_1yr.projection())
# create final degradation output layer (9999 is background), 0 is not
# degreaded, -1 is degraded
lp_perf_deg = ee.Image(-32768).where(obs_ratio_2.gte(0.5), 0) \
.where(obs_ratio_2.lte(0.5), -1)
return TEImage(lp_perf_deg.addBands(obs_ratio_2.multiply(10000)).addBands(units).unmask(-32768).int16(),
[BandInfo("Productivity performance (degradation)", add_to_map=True, metadata={'year_start': year_start, 'year_end': year_end}),
BandInfo("Productivity performance (ratio)", metadata={'year_start': year_start, 'year_end': year_end}),
BandInfo("Productivity performance (units)", metadata={'year_start': year_start})])