Created
March 8, 2022 09:22
-
-
Save prl900/9bb9b12ce9074647e09473fe334c1f27 to your computer and use it in GitHub Desktop.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
import rasterio as rio | |
from rasterio import features | |
from affine import Affine | |
import pandas as pd | |
import numpy as np | |
import geopandas as gpd | |
import numexpr as ne | |
# Input Albers tile | |
x = 20.0 | |
y = -30.0 | |
pix_res = 25.0 | |
epsg = 3577 | |
# 1- LC proc | |
ds = rio.open(f"biodiv_workflow/ga_ls_landcover_class_cyear_2_1-0-0_au_x{x}y{y}_2015-01-01_level4.tif") | |
lc = ds.read(1) | |
# 2- VE proc | |
gdf = gpd.GeoDataFrame.from_file("biodiv_workflow/QSC_Extracted_Data_20220307_151321841000-69916/data.gdb/") | |
gdf = gdf.to_crs(epsg) | |
trans = Affine(pix_res, 0.0, x*100000, 0.0, -1*pix_res, (y+1)*100000) | |
plant = features.rasterize(gdf[gdf['COVER']=='plantation'].geometry.values, out_shape=(4000,4000), fill=0, transform=trans, default_value=1, dtype='uint8') | |
nremn = features.rasterize(gdf[gdf['COVER']=='non-remnant'].geometry.values, out_shape=(4000,4000), fill=0, transform=trans, default_value=1, dtype='uint8') | |
ocean = features.rasterize(gdf[gdf['COVER']=='ocean'].geometry.values, out_shape=(4000,4000), fill=0, transform=trans, default_value=1, dtype='uint8') | |
ve = np.ones((4000,4000), dtype='uint8') | |
ve[plant==1] = 25 | |
ve[nremn==1] = 25 | |
ve[ocean==1] = 50 | |
# 3- LU proc | |
df = pd.read_csv("biodiv_workflow/FBA_LU_reclass.csv", index_col=0) | |
gdf = gpd.GeoDataFrame.from_file("biodiv_workflow/QLD_LANDUSE_June_2019/QLD_LANDUSE_June_2019.gdb/") | |
gdf = gdf.to_crs(epsg) | |
lut_values = df.index.unique().values | |
vals = [] | |
for idx in gdf.index: | |
val = gdf.iloc[idx]['QLUMP_Code'] | |
if val in lut_values: | |
val = df.loc[[gdf.iloc[idx]['QLUMP_Code']]]['out_code'].values[0] | |
else: | |
val = -1 | |
vals.append(val) | |
gdf['LUT'] = vals | |
vals = gdf['LUT'].unique() | |
vals = vals[vals>0] | |
lu = np.zeros((4000,4000), dtype='uint16') | |
for val in vals: | |
lu += features.rasterize(gdf[gdf['LUT']==val].geometry.values, out_shape=(4000,4000), fill=0, transform=trans, default_value=val, dtype='uint16') | |
# 4- Derive final product | |
#rem veg on grazing and other landuses except nature conservation, or native forestry | |
s1 = ne.evaluate('1*(((lu!=1100)&(lu!=1200)&(lu!=1200)&(lu!=2200)&(lu!=3600))&((ve!=25)&((ve!=99)&(ve!=0))))') | |
#grazing native pastures | |
s2 = ne.evaluate('2*((lu!=2100)&(((lc>3)&(lc<8))|((lc>16)&(lc<21))|(lc==84)|(lc==41)|((lc>44)&(lc<55)))&((ve==25)|(ve==99)|(ve==0)))') | |
s3 = ne.evaluate('3*((lu==2100)&(((lc>7)&(lc<12))|((lc>20)&(lc<25))|(lc==42)|((lc>56)&(lc<70)))&((ve==25)|(ve==99)|(ve==0)))') | |
s4 = ne.evaluate('4*((lu==2100)&(((lc>11)&(lc<14))|((lc>24)&(lc<27))|((lc>71)&(lc<78)))&((ve==25)|(ve==99)|(ve==0)))') | |
#native forestry | |
s22 = ne.evaluate('22*(((lu==1200)|(lu==2200)|(lu==3600))&((ve!=25)&(ve!=99)&(ve!=0)))') | |
... | |
plt.imshow(s22) |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment