read_data.py

import xarray as xr
import numpy as np

import ancillary_data as anc
import scene as scn

import os
from datetime import datetime as dt
from typing import Dict

_DTR = np.pi/180.
_RTD = 180./np.pi
_bad_data = -999.0


def read_data(sensor: str,
              l1b_filename: str,
              geo_filename: str) -> xr.Dataset:

    in_data = xr.open_dataset(geo_filename, group='geolocation_data')

    data = xr.open_dataset(l1b_filename, group='observation_data', decode_cf=False)

    if sensor.lower() == 'viirs':
        for band in list(data.variables):
            if 'reflectance' in data[band].long_name:
                if hasattr(data[band], 'VCST_scale_factor'):
                    scale_factor = data[band].VCST_scale_factor * data[band].bias_correction
                else:
                    scale_factor = data[band].scale_factor
                in_data[band] = (('number_of_lines', 'number_of_pixels'),
                                 data[band].values * scale_factor / np.cos(in_data.solar_zenith.values*_DTR))

            elif 'radiance' in data[band].long_name:
                in_data[band] = (('number_of_lines', 'number_of_pixels'),
                                 data[f'{band}_brightness_temperature_lut'].values[data[band].values])
            else:
                pass

    relazi = relative_azimuth_angle(in_data.sensor_azimuth.values, in_data.solar_azimuth.values)
    sunglint = sun_glint_angle(in_data.sensor_zenith.values, in_data.solar_zenith.values, relazi)
    scatt_angle = scattering_angle(in_data.solar_zenith.values, in_data.sensor_zenith.values, relazi)

    in_data['relative_azimuth'] = (('number_of_lines', 'number_of_pixels'), relazi)
    in_data['sunglint_angle'] = (('number_of_lines', 'number_of_pixels'), sunglint)
    in_data['scattering_angle'] = (('number_of_lines', 'number_of_pixels'), scatt_angle)

    return in_data


def relative_azimuth_angle(sensor_azimuth: np.ndarray,
                           solar_azimuth: np.ndarray) -> np.ndarray:
    rel_azimuth = np.abs(180. - np.abs(sensor_azimuth - solar_azimuth))
    return rel_azimuth


def sun_glint_angle(sensor_zenith: np.ndarray,
                    solar_zenith: np.ndarray,
                    rel_azimuth: np.ndarray) -> np.ndarray:
    cossna = (np.sin(sensor_zenith*_DTR) * np.sin(solar_zenith*_DTR) * np.cos(rel_azimuth*_DTR) +
              np.cos(sensor_zenith*_DTR) * np.cos(solar_zenith*_DTR))
    cossna[cossna > 1] = 1
    sunglint_angle = np.arccos(cossna) * _RTD
    return sunglint_angle


def scattering_angle(solar_zenith: np.ndarray,
                     sensor_zenith: np.ndarray,
                     relative_azimuth: np.ndarray) -> np.ndarray:
    cos_scatt_angle = -1. * (np.cos(solar_zenith*_DTR) * np.cos(sensor_zenith*_DTR) -
                             np.sin(solar_zenith*_DTR) * np.sin(sensor_zenith*_DTR) *
                             np.cos(relative_azimuth*_DTR))
    scatt_angle = np.arccos(cos_scatt_angle) * _RTD
    return scatt_angle


def correct_reflectances():
    pass


def read_ancillary_data(file_names: str,
                        viirs_data: xr.Dataset,
                        resolution: int = 1) -> xr.Dataset:

    # Ancillary files temporarily defined here. Eventually we will find a better way to pass these
    anc_dir = file_names['ANC_DIR']
    sst_file = file_names['SST']
    ndvi_file = file_names['NDVI']
    geos_cnst = file_names['GEOS_constants']
    geos_lnd = file_names['GEOS_land']
    geos_ocn = file_names['GEOS_ocean']
    geos1 = file_names['GEOS_atm_1']
    geos2 = file_names['GEOS_atm_2']
    vnptime = '.'.join(os.path.basename(file_names['MOD02']).split('.')[1:3])
    start_time = dt.strftime(dt.strptime(vnptime, 'A%Y%j.%H%M'), '%Y-%m-%d %H:%M:%S.000')
    # start_time = '2022-06-22 14:54:00.000'

    dim1 = viirs_data.latitude.shape[0]
    dim2 = viirs_data.latitude.shape[1]

    sst = np.empty((dim1*dim2, ), dtype=np.float32)
    ndvi = np.empty((dim1*dim2, ), dtype=np.float32)
    eco = np.empty((dim1*dim2, ), dtype=np.ubyte)

    geos_data = {'tpw': np.empty((dim1*dim2, ), dtype=np.float32),
                 'snowfr': np.empty((dim1*dim2, ), dtype=np.float32),
                 'icefr': np.empty((dim1*dim2, ), dtype=np.float32),
                 'ocnfr': np.empty((dim1*dim2, ), dtype=np.float32),
                 'landicefr': np.empty((dim1*dim2, ), dtype=np.float32),
                 'sfct': np.empty((dim1*dim2, ), dtype=np.float32),
                 }

    sst = anc.py_get_Reynolds_SST(viirs_data.latitude.values.reshape((dim1*dim2, )),
                                  viirs_data.longitude.values.reshape((dim1*dim2, )),
                                  resolution, anc_dir, sst_file, sst)
    print('SST read successfully')

    ndvi = anc.py_get_NDVI_background(viirs_data.latitude.values.reshape((dim1*dim2, )),
                                      viirs_data.longitude.values.reshape((dim1*dim2, )),
                                      resolution, anc_dir, ndvi_file, ndvi)
    print('NDVI read successfully')

    eco = anc.py_get_Olson_eco(viirs_data.latitude.values.reshape((dim1*dim2, )),
                               viirs_data.longitude.values.reshape((dim1*dim2, )),
                               resolution, anc_dir, eco)
    print('Olson eco read successfully')

    geos_data = anc.py_get_GEOS(viirs_data.latitude.values.reshape((dim1*dim2, )),
                                viirs_data.longitude.values.reshape((dim1*dim2, )),
                                resolution, start_time, anc_dir, geos1, geos2, geos_lnd, geos_ocn, geos_cnst, geos_data)
    print('GEOS5 data read successfully')

    data = viirs_data
    data['sst'] = (('number_of_lines', 'number_of_pixels'), sst.reshape((dim1, dim2)))
    data['ndvi'] = (('number_of_lines', 'number_of_pixels'), ndvi.reshape((dim1, dim2)))
    data['eco'] = (('number_of_lines', 'number_of_pixels'), eco.reshape((dim1, dim2)))
    data['geos_tpw'] = (('number_of_lines', 'number_of_pixels'), geos_data['tpw'].reshape((dim1, dim2)))
    data['geos_snowfr'] = (('number_of_lines', 'number_of_pixels'), geos_data['snowfr'].reshape((dim1, dim2)))
    data['geos_icefr'] = (('number_of_lines', 'number_of_pixels'), geos_data['icefr'].reshape((dim1, dim2)))
    data['geos_ocnfr'] = (('number_of_lines', 'number_of_pixels'), geos_data['ocnfr'].reshape((dim1, dim2)))
    data['geos_landicefr'] = (('number_of_lines', 'number_of_pixels'), geos_data['landicefr'].reshape((dim1, dim2)))
    data['geos_sfct'] = (('number_of_lines', 'number_of_pixels'), geos_data['sfct'].reshape((dim1, dim2)))

    return data


def get_data(file_names: Dict[str, str],
             sunglint_angle: float,
             hires: bool = False) -> xr.Dataset:

    mod02 = file_names['MOD02']
    mod03 = file_names['MOD03']

    if hires is True:
        img02 = file_names['IMG02']
        img03 = file_names['IMG03']

    if hires is False:
        viirs_data = read_data('viirs', f'{mod02}', f'{mod03}')
        viirs_data = read_ancillary_data(file_names, viirs_data)

        if (('M05' in viirs_data) and ('M07' in viirs_data)):
            m01 = viirs_data.M05.values
            m02 = viirs_data.M07.values
            r1 = 2.0 * (np.power(m02, 2.0) - np.power(m01, 2.0)) + (1.5 * m02) + (0.5 * m01)
            r2 = m02 + m01 + 0.5
            r3 = r1 / r2
            gemi = r3 * (1.0 - 0.25*r3) - ((m01 - 0.125) / (1.0 - m01))
        else:
            gemi = np.full((viirs_data.M15.shape), _bad_data)

        if 'M05' in viirs_data:
            idx = np.nonzero((viirs_data.M05.values < -99) | (viirs_data.M05.values > 2))
            viirs_data['M05'].values[idx] = _bad_data
        else:
            viirs_data['M05'] = (('number_of_lines', 'number_of_pixels'),
                                 np.full(viirs_data.M15.shape, _bad_data))
        if 'M07' in viirs_data:
            idx = np.nonzero((viirs_data.M07.values < -99) | (viirs_data.M07.values > 2))
            viirs_data['M07'].values[idx] = _bad_data
        else:
            viirs_data['M07'] = (('number_of_lines', 'number_of_pixels'),
                                 np.full(viirs_data.M15.shape, _bad_data))

        idx = np.nonzero((viirs_data.M12.values < 0) | (viirs_data.M12.values > 1000))
        viirs_data['M12'].values[idx] = _bad_data
        idx = np.nonzero((viirs_data.M13.values < 0) | (viirs_data.M13.values > 1000))
        viirs_data['M13'].values[idx] = _bad_data
        idx = np.nonzero((viirs_data.M14.values < 0) | (viirs_data.M14.values > 1000))
        viirs_data['M14'].values[idx] = _bad_data
        idx = np.nonzero((viirs_data.M15.values < 0) | (viirs_data.M15.values > 1000))
        viirs_data['M15'].values[idx] = _bad_data
        idx = np.nonzero((viirs_data.M16.values < 0) | (viirs_data.M16.values > 1000))
        viirs_data['M16'].values[idx] = _bad_data

        # Compute channel differences and ratios that are used in the tests
        viirs_data['M14-M15'] = (('number_of_lines', 'number_of_pixels'),
                                 viirs_data.M14.values - viirs_data.M15.values)
        viirs_data['M15-M16'] = (('number_of_lines', 'number_of_pixels'),
                                 viirs_data.M15.values - viirs_data.M16.values)
        viirs_data['M15-M12'] = (('number_of_lines', 'number_of_pixels'),
                                 viirs_data.M15.values - viirs_data.M12.values)
        viirs_data['M13-M16'] = (('number_of_lines', 'number_of_pixels'),
                                 viirs_data.M13.values - viirs_data.M16.values)
        viirs_data['M07-M05ratio'] = (('number_of_lines', 'number_of_pixels'),
                                      viirs_data.M07.values / viirs_data.M05.values)
        viirs_data['GEMI'] = (('number_of_lines', 'number_of_pixels'), gemi)

        # temp value to force the code to work
        viirs_data['M128'] = (('number_of_lines', 'number_of_pixels'), np.zeros(viirs_data.M15.shape))

    else:
        viirs_data = read_data('viirs', f'{img02}', f'{img03}')
        viirs_data['M05'] = viirs_data.I01
        viirs_data['M07'] = viirs_data.I02

        idx = np.nonzero((viirs_data.M05.values < -99) | (viirs_data.M05.values > 2))
        viirs_data['M05'].values[idx] = _bad_data
        idx = np.nonzero((viirs_data.M07.values < -99) | (viirs_data.M07.values > 2))
        viirs_data['M07'].values[idx] = _bad_data

        idx = np.nonzero((viirs_data.I01.values < 0) | (viirs_data.I01.values > 1000))
        viirs_data['I01'].values[idx] = _bad_data
        idx = np.nonzero((viirs_data.I02.values < 0) | (viirs_data.I02.values > 1000))
        viirs_data['I02'].values[idx] = _bad_data
        idx = np.nonzero((viirs_data.I03.values < 0) | (viirs_data.I03.values > 1000))
        viirs_data['I03'].values[idx] = _bad_data
        idx = np.nonzero((viirs_data.I04.values < 0) | (viirs_data.I04.values > 1000))
        viirs_data['I04'].values[idx] = _bad_data
        idx = np.nonzero((viirs_data.I05.values < 0) | (viirs_data.I05.values > 1000))
        viirs_data['I05'].values[idx] = _bad_data

        viirs_data = read_ancillary_data(file_names, viirs_data, resolution=2)

        viirs_data['I05-I04'] = (('number_of_lines_2', 'number_of_pixels_2'),
                                 viirs_data.I05.values - viirs_data.I04.values)
        viirs_data['I02-I01ratio'] = (('number_of_lines_2', 'number_of_pixels_2'),
                                      viirs_data.I02.values / viirs_data.I01.values)

    scene_flags = scn.find_scene(viirs_data, sunglint_angle)
    scene = scn.scene_id(scene_flags)

    scene_xr = xr.Dataset()
    for s in scn._scene_list:
        scene_xr[s] = (('number_of_lines', 'number_of_pixels'), scene[s])

    scene['lat'] = viirs_data.latitude
    scene['lon'] = viirs_data.longitude

    data = xr.Dataset(viirs_data, coords=scene_xr)
    data.drop_vars(['latitude', 'longitude'])

    return data