Annual runoff in 1-degree latitude bins, split E/W

[mankoff]

Issues:

• The E/W split fails at the northern and southern extremes, but for most of Greenland a simple E/W split works reasonably well.
• Splitting the ice runoff into submarine discharge and ice that flows out via terrestrial streams is done with a simple altitude < or > 0 threshold. The altitude metadata has some quality issues (see discussion in paper). A better method might use the distance from the outlet to the coastline, and assume all outlets < X km from the coast are really submarine discharge.

This code generates CSVs for all land or ice runoff within 1 degree lat bins:

import xarray as xr

type = 'ice' # 'ice' or 'land'
type = 'land'
ds = xr.open_mfdataset('/home/kdm/data/Mankoff_2020/water/'+type+'/runoff/RACMO*.nc', combine='by_coords')

ds = ds.resample({'time':'YS'}).sum()
ds['lat'] = np.floor(ds['lat']).astype(np.int)

dfE = ds.where(ds['lon'] < -44, drop=True).groupby("lat").sum().to_dataframe()['runoff']
dfW = ds.where(ds['lon'] >= -44, drop=True).groupby("lat").sum().to_dataframe()['runoff']

lats = dfE.index.get_level_values('lat').unique().astype(np.int)
for lat in lats:
    print(lat)
    dfE.loc[lat,:].droplevel(0).to_csv(f'./runoff_1deg/{type}_E_{lat}_{lat+1}.csv')
    dfW.loc[lat,:].droplevel(0).to_csv(f'./runoff_1deg/{type}_W_{lat}_{lat+1}.csv')

This code splits the ice runoff into terrestrial vs submarine:

import xarray as xr

type = 'terrestrial' # 'submarine' or 'terrestrial'
# type = 'submarine'
assert((type == 'submarine') | (type == 'terrestrial'))

ds = xr.open_mfdataset('/home/kdm/data/Mankoff_2020/water/ice/runoff/RACMO*.nc', combine='by_coords')

ds = ds.resample({'time':'YS'}).sum()
ds['lat'] = np.floor(ds['lat']).astype(np.int)

if type == 'submarine':
    ds = ds.where(ds['alt'] < 0, drop=True)
elif type == 'terrestrial':
    ds = ds.where(ds['alt'] >= 0, drop=True)
    
dfE = ds.where(ds['lon'] < -44, drop=True).groupby("lat").sum().to_dataframe()['runoff']
dfW = ds.where(ds['lon'] >= -44, drop=True).groupby("lat").sum().to_dataframe()['runoff']

lats = dfE.index.get_level_values('lat').unique().astype(np.int)
for lat in lats:
    print(lat)
    dfE.loc[lat,:].droplevel(0).to_csv(f'./runoff_1deg/{type}_ice_E_{lat}_{lat+1}.csv')
    dfW.loc[lat,:].droplevel(0).to_csv(f'./runoff_1deg/{type}_ice_W_{lat}_{lat+1}.csv')