STIC-JPL 1.1.0__py3-none-any.whl → 1.2.2__py3-none-any.whl

STIC_JPL/FVC_from_NDVI.py

@@ -0,0 +1,49 @@
+from typing import Union
+import numpy as np
+import rasters as rt
+from rasters import Raster
+
+KPAR = 0.5
+MIN_FIPAR = 0.0
+MAX_FIPAR = 1.0
+MIN_LAI = 0.0
+MAX_LAI = 10.0
+
+def FVC_from_NDVI(NDVI: Union[Raster, np.ndarray]) -> Union[Raster, np.ndarray]:
+    """
+    Estimate Fractional Vegetation Cover (FVC) from Normalized Difference Vegetation Index (NDVI)
+    using a scaled NDVI approach.
+
+    This method linearly scales NDVI values between two endmembers:
+        - NDVIs: NDVI value for bare soil (typically ~0.04 ± 0.03)
+        - NDVIv: NDVI value for full vegetation (typically ~0.52 ± 0.03)
+
+    The resulting Fractional Vegetation Cover (FVC) is calculated as:
+
+        FVC = clip((NDVI - NDVIs) / (NDVIv - NDVIs), 0.0, 1.0)
+
+    This approach is based on the assumption that NDVI increases linearly with vegetation cover
+    between these two extremes, and is well-supported in the literature.
+
+    References:
+        Carlson, T. N., & Ripley, D. A. (1997). On the relation between NDVI, fractional vegetation cover,
+        and leaf area index. Remote Sensing of Environment, 62(3), 241–252.
+        https://doi.org/10.1016/S0034-4257(97)00104-1
+
+        Gutman, G., & Ignatov, A. (1998). The derivation of the green vegetation fraction from NOAA/AVHRR
+        data for use in numerical weather prediction models. International Journal of Remote Sensing,
+        19(8), 1533–1543. https://doi.org/10.1080/014311698215333
+
+    Parameters:
+        NDVI (Union[Raster, np.ndarray]): Input NDVI data.
+
+    Returns:
+        Union[Raster, np.ndarray]: Estimated Fractional Vegetation Cover (FVC).
+    """
+    NDVIv = 0.52  # NDVI for fully vegetated pixel
+    NDVIs = 0.04  # NDVI for bare soil pixel
+
+    # Scale NDVI to FVC using a linear model and clip to [0, 1]
+    FVC = rt.clip((NDVI - NDVIs) / (NDVIv - NDVIs), 0.0, 1.0)
+
+    return FVC

STIC_JPL/LAI_from_NDVI.py

@@ -0,0 +1,61 @@
+from typing import Union
+import numpy as np
+import rasters as rt
+from rasters import Raster
+
+# Constants
+KPAR = 0.5  # Extinction coefficient for PAR, assumed average for broadleaf canopies (Weiss & Baret, 2010)
+MIN_FIPAR = 0.0
+MAX_FIPAR = 1.0
+MIN_LAI = 0.0
+MAX_LAI = 10.0
+
+def LAI_from_NDVI(
+        NDVI: Union[Raster, np.ndarray],
+        min_fIPAR: float = MIN_FIPAR,
+        max_fIPAR: float = MAX_FIPAR,
+        min_LAI: float = MIN_LAI,
+        max_LAI: float = MAX_LAI) -> Union[Raster, np.ndarray]:
+    """
+    Estimate Leaf Area Index (LAI) from NDVI using a simplified two-step empirical model.
+
+    This method first approximates the fraction of absorbed photosynthetically active radiation (fIPAR)
+    from NDVI, and then estimates LAI using the Beer–Lambert Law. The extinction coefficient for PAR (KPAR)
+    is assumed to be 0.5, which is typical for broadleaf canopies under diffuse light conditions.
+
+    Steps:
+    1. fIPAR ≈ NDVI - 0.05 (empirical offset to account for soil background and sensor noise)
+       - Based on observed relationships in Myneni & Williams (1994)
+    2. LAI = -ln(1 - fIPAR) / KPAR (Beer–Lambert Law)
+       - From Sellers (1985)
+
+    All outputs are clipped to user-defined minimum and maximum thresholds to ensure physical realism.
+
+    Parameters:
+        NDVI (Union[Raster, np.ndarray]): Input NDVI data.
+        min_fIPAR (float): Minimum fIPAR value for clipping (default 0.0).
+        max_fIPAR (float): Maximum fIPAR value for clipping (default 1.0).
+        min_LAI (float): Minimum LAI value for clipping (default 0.0).
+        max_LAI (float): Maximum LAI value for clipping (default 10.0).
+
+    Returns:
+        Union[Raster, np.ndarray]: Estimated LAI values.
+
+    References:
+        - Sellers, P. J. (1985). Canopy reflectance, photosynthesis and transpiration. 
+          *International Journal of Remote Sensing*, 6(8), 1335–1372.
+        - Myneni, R. B., & Williams, D. L. (1994). On the relationship between FAPAR and NDVI. 
+          *Remote Sensing of Environment*, 49(3), 200–211.
+        - Weiss, M., & Baret, F. (2010). CAN-EYE V6.1 User Manual. INRA-CSE.
+
+    """
+    # Empirical conversion from NDVI to fIPAR (adjusted for background noise)
+    fIPAR = rt.clip(NDVI - 0.05, min_fIPAR, max_fIPAR)
+
+    # Avoid division by zero or log of 0 by masking zero fIPAR values
+    fIPAR = np.where(fIPAR == 0, np.nan, fIPAR)
+
+    # Apply Beer–Lambert law to estimate LAI
+    LAI = rt.clip(-np.log(1 - fIPAR) * (1 / KPAR), min_LAI, max_LAI)
+
+    return LAI

STIC_JPL/STIC_JPL.py

@@ -1,370 +1,3 @@
-from typing import Union, Callable
-import logging
-from datetime import datetime, timedelta
-from os.path import join, abspath, expanduser
-from typing import Dict, List
-import numpy as np
-import warnings
-from .diagnostic import diagnostic
-import colored_logging as cl
-from .meteorology_conversion import calculate_air_density, calculate_specific_heat, calculate_specific_humidity, calculate_surface_pressure, celcius_to_kelvin
-import rasters as rt
-from GEOS5FP import GEOS5FP
-from solar_apparent_time import solar_day_of_year_for_area, solar_hour_of_day_for_area
-
-from .timer import Timer
-
-from rasters import Raster, RasterGeometry
-
-from .vegetation_conversion.vegetation_conversion import FVC_from_NDVI, LAI_from_NDVI
-
-from .constants import *
-from .closure import STIC_closure
-from .soil_moisture_initialization import initialize_soil_moisture
-from .soil_moisture_iteration import iterate_soil_moisture
-from .net_radiation import calculate_net_longwave_radiation
-from .initialize_with_solar import initialize_with_solar
-from .canopy_air_stream import calculate_canopy_air_stream_vapor_pressure
-from .initialize_without_solar import initialize_without_solar
-from .iterate_with_solar import iterate_with_solar
-from .iterate_without_solar import iterate_without_solar
-from .root_zone_initialization import calculate_root_zone_moisture
-
-from .soil_heat_flux import calculate_SEBAL_soil_heat_flux
-
-__author__ = 'Kaniska Mallick, Madeleine Pascolini-Campbell, Gregory Halverson'
-
-logger = logging.getLogger(__name__)
-
-def STIC_JPL(
-        ST_C: Union[Raster, np.ndarray],
-        emissivity: Union[Raster, np.ndarray],
-        NDVI: Union[Raster, np.ndarray],
-        albedo: Union[Raster, np.ndarray],
-        Rn_Wm2: Union[Raster, np.ndarray],
-        geometry: RasterGeometry = None,
-        time_UTC: datetime = None,
-        hour_of_day: np.ndarray = None,
-        day_of_year: np.ndarray = None,
-        GEOS5FP_connection: GEOS5FP = None,
-        Ta_C: Union[Raster, np.ndarray] = None,
-        RH: Union[Raster, np.ndarray] = None,
-        G: Union[Raster, np.ndarray] = None,
-        G_method: str = DEFAULT_G_METHOD,
-        SM: Union[Raster, np.ndarray] = None,
-        Rg_Wm2: Union[Raster, np.ndarray] = None,
-        FVC: Union[Raster, np.ndarray] = None,
-        LAI: Union[Raster, np.ndarray] = None,
-        elevation_m: Union[Raster, np.ndarray] = None,
-        delta_hPa: Union[Raster, np.ndarray] = None,
-        gamma_hPa: Union[Raster, np.ndarray, float] = GAMMA_HPA,
-        rho_kgm3: Union[Raster, np.ndarray] = RHO_KGM3,
-        Cp_Jkg: Union[Raster, np.ndarray] = CP_JKG,
-        alpha: float = PT_ALPHA,
-        LE_convergence_target: float = LE_CONVERGENCE_TARGET_WM2,
-        max_iterations: int = MAX_ITERATIONS,
-        diagnostic_directory: str = None,
-        show_distributions: bool = SHOW_DISTRIBUTIONS,
-        use_variable_alpha: bool = USE_VARIABLE_ALPHA) -> Dict[str, Union[Raster, np.ndarray]]:
-    results = {}
-
-    if geometry is None and isinstance(ST_C, Raster):
-        geometry = ST_C.geometry
-
-    if GEOS5FP_connection is None:
-        GEOS5FP_connection = GEOS5FP()
-
-    if (day_of_year is None or hour_of_day is None) and time_UTC is not None and geometry is not None:
-        day_of_year = solar_day_of_year_for_area(time_UTC=time_UTC, geometry=geometry)
-        hour_of_day = solar_hour_of_day_for_area(time_UTC=time_UTC, geometry=geometry)
-
-    if time_UTC is None and day_of_year is None and hour_of_day is None:
-        raise ValueError("no time given between time_UTC, day_of_year, and hour_of_day")
-
-    diag_kwargs = {
-        "show_distributions": show_distributions, 
-        "output_directory": diagnostic_directory
-    }
-
-    seconds_of_day = hour_of_day * 3600.0
-
-    # load air temperature in Celsius if not provided
-    if Ta_C is None:
-        Ta_C = GEOS5FP_connection.Ta_C(time_UTC=time_UTC, geometry=geometry, resampling=resampling)
-
-    # load relative humidity if not provided
-    if RH is None:
-        RH = GEOS5FP_connection.RH(time_UTC=time_UTC, geometry=geometry, resampling=resampling)
-
-    # calculate fraction of vegetation cover if it's not given
-    if FVC is None:
-        FVC = FVC_from_NDVI(NDVI)
-    
-    # calculate leaf area index if it's not given
-    if LAI is None:
-        LAI = LAI_from_NDVI(NDVI)
-
-    # saturation air pressure in hPa
-    SVP_hPa = 6.13753 * (np.exp((17.27 * Ta_C) / (Ta_C + 237.3)))
-
-    # calculate delta term if it's not given
-    if delta_hPa is None:
-        # slope of saturation vapor pressure to air temperature (hpa/K)
-        delta_hPa = 4098 * SVP_hPa / (Ta_C + 237.3) ** 2
-
-    Ta_K = celcius_to_kelvin(Ta_C)
-
-    # actual vapor pressure at TA (hpa/K)
-    Ea_hPa = SVP_hPa * (RH)
-    Ea_Pa = Ea_hPa * 100.0
-
-    # vapor pressure deficit (hPa)
-    VPD_hPa = SVP_hPa - Ea_hPa
-
-    # swapping in the dew-point calculation from PT-JPL
-    Td_C = Ta_C - ((100 - RH * 100) / 5.0)
-
-    # difference between surface and air temperature (Celsius)
-    dTS_C = ST_C - Ta_C
-
-    # saturation vapor pressure at surface temperature (hPa/K)
-    Estar_hPa = 6.13753 * np.exp((17.27 * ST_C) / (ST_C + 237.3))
-
-    if Rg_Wm2 is None:
-        # if G is None and SM is None:
-        #     raise ValueError("soil heat flux or soil moisture prior required if solar radiation is not given")
-
-        if G is None:
-            G = calculate_SEBAL_soil_heat_flux(
-                ST_C=ST_C,
-                NDVI=NDVI,
-                albedo=albedo,
-                Rn=Rn_Wm2,
-            )
-        
-        phi_Wm2 = Rn_Wm2 - G
-
-        # initialize without solar radiation
-        SM, SMrz, s1, s3, s33, s44, Ms, Tsd_C, Es_hPa, Ds = initialize_without_solar(
-            ST_C = ST_C,  # Surface temperature in Celsius
-            Ta_C = Ta_C,  # Air temperature in Celsius
-            dTS = dTS_C,  # Temperature difference between surface and air in Celsius
-            Td_C = Td_C,  # Dewpoint temperature in Celsius
-            Ea_hPa = Ea_hPa,  # Actual vapor pressure in hPa
-            Estar_hPa = Estar_hPa,  # Saturation vapor pressure at surface temperature (hPa/K)
-            SVP_hPa = SVP_hPa,  # Saturation vapor pressure at the surface in hPa
-            delta_hPa = delta_hPa,  # Slope of the saturation vapor pressure-temperature curve in hPa/K
-            phi_Wm2 = phi_Wm2,  # Available energy in W/m2
-            gamma_hPa = gamma_hPa,  # Psychrometric constant in hPa/°C
-            alpha = alpha  # Priestley-Taylor alpha
-        )
-    else:
-        SM, SMrz, Ms, s1, s3, Ep_PT, Rnsoil, LWnet_Wm2, G, Tsd_C, Ds, Es_hPa, phi_Wm2 = initialize_with_solar(
-            seconds_of_day = seconds_of_day,  # time of day in seconds since midnight
-            Rg_Wm2 = Rg_Wm2,  # solar radiation (W/m^2)
-            Rn_Wm2 = Rn_Wm2,  # net radiation (W/m^2)
-            ST_C = ST_C,  # surface temperature (Celsius)
-            emissivity = emissivity,  # emissivity of the surface
-            Ta_C = Ta_C,  # air temperature (Celsius)
-            dTS_C = dTS_C,  # surface air temperature difference (Celsius)
-            Td_C = Td_C,  # dew point temperature (Celsius)
-            VPD_hPa = VPD_hPa,  # vapor pressure deficit (hPa)
-            SVP_hPa = SVP_hPa,  # saturation vapor pressure at given air temperature (hPa)
-            Ea_hPa = Ea_hPa,  # actual vapor pressure at air temperature (hPa)
-            Estar_hPa = Estar_hPa,  # saturation vapor pressure at surface temperature (hPa)
-            delta_hPa = delta_hPa,  # slope of saturation vapor pressure to air temperature (hpa/K)
-            NDVI=NDVI,  # normalized difference vegetation index
-            FVC = FVC,  # fractional vegetation cover
-            LAI = LAI,  # leaf area index
-            albedo = albedo,  # albedo of the surface
-            gamma_hPa=gamma_hPa,  # psychrometric constant (hPa/°C)
-            G_method = DEFAULT_G_METHOD,  # method for calculating soil heat flux
-        )
-    
-    diagnostic(Ms, "Ms", **diag_kwargs)
-
-    # STIC analytical equations (convergence on LE)
-    gB_ms, gS_ms, dT_C, EF = STIC_closure(
-        delta_hPa=delta_hPa, 
-        phi_Wm2=phi_Wm2, 
-        Es_hPa=Es_hPa, 
-        Ea_hPa=Ea_hPa, 
-        Estar_hPa=Estar_hPa, 
-        SM=SM,
-        gamma_hPa=gamma_hPa,
-        rho_kgm3=rho_kgm3,
-        Cp_Jkg=Cp_Jkg,
-        alpha=alpha
-    )
-    
-    gBB = gB_ms
-    gSS = gS_ms
-    gBB_by_gSS = rt.where(gSS == 0, 0, gBB / gSS)
-    gB_by_gS = rt.where(gS_ms == 0, 0, gB_ms / gS_ms)
-    dT_C = dT_C
-    T0_C = dT_C + Ta_C
-    
-    PET_Wm2 = ((delta_hPa * phi_Wm2 + rho_kgm3 * Cp_Jkg * gB_ms * VPD_hPa) / (delta_hPa + gamma_hPa))  # Penman potential evaporation
-    
-    gR = (4 * SB_SIGMA * (Ta_C + 273) ** 3 * emissivity) / (rho_kgm3 * Cp_Jkg)
-    omega = ((delta_hPa / gamma_hPa) + 1) / ((delta_hPa / gamma_hPa) + 1 + gB_by_gS)
-    LE_eq = (phi_Wm2 * (delta_hPa / gamma_hPa)) / ((delta_hPa / gamma_hPa) + 1)
-    LE_imp = (Cp_Jkg * 0.0289644 / gamma_hPa) * gS_ms * 40 * VPD_hPa
-    LE_init = omega * LE_eq + (1 - omega) * LE_imp
-    dry = (Ds > VPD_hPa) & (PET_Wm2 > phi_Wm2) & (dTS_C > 0) & (Td_C <= 0)
-    omega = rt.where(dry,
-                        ((delta_hPa / gamma_hPa) + 1 + gR / gB_ms) / ((delta_hPa / gamma_hPa) + 1 + gB_ms / gS_ms + gR / gS_ms + gR / gB_ms),
-                        omega)
-    LE_eq = rt.where(dry, (phi_Wm2 * (delta_hPa / gamma_hPa)) / ((delta_hPa / gamma_hPa) + 1 + gR / gB_ms), LE_eq)
-    LE_init = rt.where(dry, omega * LE_eq + (1 - omega * LE_imp), LE_init)
-    
-    # sensible heat flux
-    H_Wm2 = ((gamma_hPa * phi_Wm2 * (1 + gB_by_gS) - rho_kgm3 * Cp_Jkg * gB_ms * VPD_hPa) / (delta_hPa + gamma_hPa * (1 + (gB_by_gS))))
-    
-    LE_Wm2_new = LE_init
-    LE_Wm2_change = LE_convergence_target
-    LE_Wm2_old = LE_Wm2_new
-    LE_transpiration_Wm2 = None
-    PT_Wm2 = None
-    iteration = 1
-    LE_Wm2_max_change = 0
-    t = Timer()
-
-    while (np.nanmax(LE_Wm2_change) >= LE_convergence_target and iteration <= max_iterations):
-        logger.info(f"running STIC iteration {cl.val(iteration)} / {cl.val(max_iterations)}")
-
-        if Rg_Wm2 is None:
-            SM, SMrz, Ms, s1, e0, e0star, Tsd_C, D0, alphaN = iterate_without_solar(
-                LE = LE_Wm2_new,  # Latent heat flux (W/m^2)
-                PET = PET_Wm2,  # Potential evapotranspiration (W/m^2)
-                SM = SM,
-                ST_C = ST_C,  # Surface temperature (°C)
-                Ta_C = Ta_C,  # Air temperature (°C)
-                dTS = dTS_C,  # Surface-air temperature difference (°C)
-                T0 = T0_C,  # Reference temperature (°C)
-                gB = gB_ms,  # Boundary layer conductance (m/s)
-                gS = gS_ms,  # Stomatal conductance (m/s)
-                Ea_hPa = Ea_hPa,  # Actual vapor pressure (hPa)
-                Td_C = Td_C,  # Dew point temperature (°C)
-                VPD_hPa = VPD_hPa,  # Vapor pressure deficit (hPa)
-                Estar = Estar_hPa,  # Saturation vapor pressure at surface temperature (hPa)
-                delta = delta_hPa,  # Slope of the saturation vapor pressure-temperature curve (hPa/°C)
-                phi = phi_Wm2,  # available energy (W/m^2)
-                Ds = Ds,  # Vapor pressure deficit at source (hPa)
-                Es = Es_hPa,  # Saturation vapor pressure (hPa)
-                s3 = s3,  # Slope of the saturation vapor pressure and temperature 
-                s4 = s44,  # Slope of the saturation vapor pressure and temperature 
-                gB_by_gS = gB_by_gS,  # Ratio of boundary layer conductance to stomatal conductance
-                gamma_hPa = gamma_hPa,  # Psychrometric constant (hPa/°C)
-                rho_kgm3 = rho_kgm3,  # Air density (kg/m^3)
-                Cp_Jkg = Cp_Jkg  # Specific heat at constant pressure (J/kg/K)
-            )
-        else:
-            SM, G, e0, e0star, D0, alphaN = iterate_with_solar(
-                seconds_of_day = seconds_of_day,  # Seconds of the day
-                ST_C = ST_C,  # Soil temperature (°C)
-                NDVI = NDVI,  # Normalized Difference Vegetation Index
-                albedo = albedo,  # Albedo
-                gB_ms = gB_ms,  # boundary layer conductance (m/s)
-                gS_ms = gS_ms,  # stomatal conductance (m/s)
-                LE_Wm2 = LE_Wm2_new,  # latent heat flux (W/m^2)
-                Rg_Wm2 = Rg_Wm2,  # Incoming solar radiation (W/m^2)
-                Rn_Wm2 = Rn_Wm2,  # Net radiation (W/m^2)
-                LWnet_Wm2 = LWnet_Wm2,  # Net longwave radiation (W/m^2)
-                Ta_C = Ta_C,  # Air temperature (°C)
-                dTS_C = dTS_C,  # Change in soil temperature (°C)
-                Td_C = Td_C,  # Dew point temperature (°C)
-                Tsd_C = Tsd_C,  # Soil dew point temperature (°C)
-                Ea_hPa = Ea_hPa,  # actual vapor pressure (hPa)
-                Estar_hPa = Estar_hPa,  # saturation vapor pressure at surface temperature (hPa)
-                VPD_hPa = VPD_hPa,  # Vapor pressure deficit (hPa)
-                SVP_hPa = SVP_hPa,  # Saturation vapor pressure (hPa)
-                delta_hPa = delta_hPa,  # Slope of the saturation vapor pressure-temperature curve (hPa/°C)
-                phi_Wm2 = phi_Wm2,  # Net radiation minus soil heat flux (W/m^2)
-                Es_hPa = Es_hPa,  # Saturation vapor pressure (hPa)
-                s1 = s1,  # Soil moisture parameter
-                s3 = s3,  # Soil moisture parameter
-                FVC = FVC,  # Fractional canopy cover
-                T0_C = T0_C,  # Reference temperature (°C)
-                gB_by_gS = gB_by_gS,  # Ratio of boundary layer conductance to stomatal conductance
-                gamma_hPa = gamma_hPa,  # Psychrometric constant (hPa/°C)
-                rho_kgm3 = rho_kgm3,  # Air density (kg/m^3)
-                Cp_Jkg = Cp_Jkg,  # Specific heat at constant pressure (J/kg/K)
-                G_method = "santanello"  # Method for calculating soil heat flux
-            )
-
-        if use_variable_alpha:
-            alpha = alphaN
-            logger.info(f"using variable Priestley-Taylor alpha with mean: {cl.val(np.round(np.nanmean(alpha), 3))}")
-
-        # re-estimated conductances and states
-        gB_ms, gS_ms, dT_C, EF = STIC_closure(
-            delta_hPa=delta_hPa,  # Slope of the saturation vapor pressure-temperature curve (hPa/°C)
-            phi_Wm2=phi_Wm2,  # available energy (W/m^2)
-            Es_hPa=Es_hPa,  # Vapor pressure at the reference height (hPa)
-            Ea_hPa=Ea_hPa,  # Actual vapor pressure (hPa)
-            Estar_hPa=Estar_hPa,  # Saturation vapor pressure at the reference height (hPa)
-            SM=SM,  # Soil moisture (m³/m³)
-            gamma_hPa=gamma_hPa,  # Psychrometric constant (hPa/°C)
-            rho_kgm3=rho_kgm3,  # Air density (kg/m³)
-            Cp_Jkg=Cp_Jkg,  # Specific heat capacity of air (J/kg/°C)
-            alpha=alpha  # Stability correction factor for conductance 
-        )
-
-        gB_by_gS = rt.where(gS_ms == 0, 0, gB_ms / gS_ms)
-        T0_C = dT_C + Ta_C
-        # latent heat flux
-        LE_Wm2_new = ((delta_hPa * phi_Wm2 + rho_kgm3 * Cp_Jkg * gB_ms * VPD_hPa) / (delta_hPa + gamma_hPa * (1 + gB_by_gS)))
-        LE_Wm2_new = rt.where(LE_Wm2_new > phi_Wm2, phi_Wm2, LE_Wm2_new)
-        # Sensible Heat Flux
-        H_Wm2 = ((gamma_hPa * phi_Wm2 * (1 + gB_by_gS) - rho_kgm3 * Cp_Jkg * gB_ms * VPD_hPa) / (delta_hPa + gamma_hPa * (1 + (gB_by_gS))))
-        # potential evaporation (Penman)
-        PET_Wm2 = ((delta_hPa * phi_Wm2 + rho_kgm3 * Cp_Jkg * gB_ms * VPD_hPa) / (delta_hPa + gamma_hPa))
-        # Potential Transpiration
-        PT_Wm2 = (delta_hPa * phi_Wm2 + rho_kgm3 * Cp_Jkg * gB_ms * VPD_hPa) / (delta_hPa + gamma_hPa * (1 + SM * gB_by_gS))  # potential transpiration
-        # ET PARTITIONING
-        LE_soil_Wm2 = rt.clip(SM * PET_Wm2, 0, None)
-        LE_transpiration_Wm2 = rt.clip(LE_Wm2_new - LE_soil_Wm2, 0, None)
-        # change in latent heat flux estimate
-        LE_Wm2_change = np.abs(LE_Wm2_old - LE_Wm2_new)
-        LE_Wm2_new = rt.where(np.isnan(LE_Wm2_new), LE_Wm2_old, LE_Wm2_new)
-        LE_Wm2_old = LE_Wm2_new
-        LE_Wm2_max_change = np.nanmax(LE_Wm2_change)
-        logger.info(
-            f"completed STIC iteration {cl.val(iteration)} / {cl.val(max_iterations)} with max LE change: {cl.val(np.round(LE_Wm2_max_change, 3))} ({t} seconds)")
-        
-        diagnostic(SM, f"SM_{iteration}", **diag_kwargs)
-        diagnostic(G, f"G_{iteration}", **diag_kwargs)
-        diagnostic(LE_Wm2_new, f"LE_{iteration}", **diag_kwargs)
-
-        if LE_Wm2_max_change <= LE_convergence_target:
-            logger.info(f"max LE change {cl.val(np.round(LE_Wm2_max_change, 3))} within convergence target {cl.val(np.round(LE_convergence_target, 3))} with {cl.val(iteration)} iteration{'s' if iteration > 1 else ''}")
-
-        iteration += 1
-
-    iteration -= 1
-    results["LE_max_change"] = LE_Wm2_max_change
-    results["iteration"] = iteration
-
-    LE = LE_Wm2_new
-
-    results["LE"] = LE
-    results["LE_change"] = LE_Wm2_change
-    results["LEt"] = LE_transpiration_Wm2
-    results["PT"] = PT_Wm2
-    results["PET"] = PET_Wm2
-    results["G"] = G
-
-    if isinstance(geometry, RasterGeometry):
-        for name, array in results.items():
-            try:
-                results[name] = Raster(array.reshape(geometry.shape), geometry=geometry)
-            except Exception as e:
-                pass
-        
-        results["LE"].cmap = ET_COLORMAP
-        results["PET"].cmap = ET_COLORMAP
-
-    warnings.resetwarnings()
-
-    return results
+from .model import *
+from .generate_STIC_inputs import generate_STIC_inputs
+from .process_STIC_table import process_STIC_table

STIC_JPL/celcius_to_kelvin.py

@@ -0,0 +1,11 @@
+from typing import Union
+import numpy as np
+from rasters import Raster
+
+def celcius_to_kelvin(T_C: Union[Raster, np.ndarray]) -> Union[Raster, np.ndarray]:
+    """
+    convert temperature in celsius to kelvin.
+    :param T_C: temperature in celsius
+    :return: temperature in kelvin
+    """
+    return T_C + 273.15

STIC_JPL/generate_STIC_inputs.py

@@ -0,0 +1,65 @@
+import logging
+
+import numpy as np
+from dateutil import parser
+from pandas import DataFrame
+from rasters import Point
+from sentinel_tiles import sentinel_tiles
+from solar_apparent_time import UTC_to_solar
+from SEBAL_soil_heat_flux import calculate_SEBAL_soil_heat_flux
+
+from .model import STIC_JPL, MAX_ITERATIONS, USE_VARIABLE_ALPHA
+
+logger = logging.getLogger(__name__)
+
+def generate_STIC_inputs(STIC_inputs_from_calval_df: DataFrame) -> DataFrame:
+    """
+    STIC_inputs_from_claval_df:
+        Pandas DataFrame containing the columns: tower, lat, lon, time_UTC, albedo, elevation_km
+    return:
+        Pandas DataFrame containing the columns: tower, lat, lon, time_UTC, doy, albedo, elevation_km, AOT, COT, vapor_gccm, ozone_cm, SZA, KG
+    """
+    # output_rows = []
+    STIC_inputs_df = STIC_inputs_from_calval_df.copy()
+
+    hour_of_day = []
+    doy = []
+    Topt = []
+    fAPARmax = []
+
+    for i, input_row in STIC_inputs_from_calval_df.iterrows():
+        tower = input_row.tower
+        lat = input_row.lat
+        lon = input_row.lon
+        time_UTC = input_row.time_UTC
+        albedo = input_row.albedo
+        elevation_km = input_row.elevation_km
+        logger.info(f"collecting STIC inputs for tower {tower} lat {lat} lon {lon} time {time_UTC} UTC")
+        time_UTC = parser.parse(str(time_UTC))
+        time_solar = UTC_to_solar(time_UTC, lon)
+        hour_of_day.append(time_solar.hour)
+        doy.append(time_UTC.timetuple().tm_yday)
+        date_UTC = time_UTC.date()
+        tile = sentinel_tiles.toMGRS(lat, lon)[:5]
+        tile_grid = sentinel_tiles.grid(tile=tile, cell_size=70)
+        rows, cols = tile_grid.shape
+        row, col = tile_grid.index_point(Point(lon, lat))
+        geometry = tile_grid[max(0, row - 1):min(row + 2, rows - 1),
+                             max(0, col - 1):min(col + 2, cols - 1)]
+    
+    if not "hour_of_day" in STIC_inputs_df.columns:
+        STIC_inputs_df["hour_of_day"] = hour_of_day
+
+    if not "doy" in STIC_inputs_df.columns:
+        STIC_inputs_df["doy"] = doy
+    
+    if not "Topt" in STIC_inputs_df.columns:
+        STIC_inputs_df["Topt"] = Topt
+    
+    if not "fAPARmax" in STIC_inputs_df.columns:
+        STIC_inputs_df["fAPARmax"] = fAPARmax
+    
+    if "Ta" in STIC_inputs_df and "Ta_C" not in STIC_inputs_df:
+        STIC_inputs_df.rename({"Ta": "Ta_C"}, inplace=True)
+    
+    return STIC_inputs_df

STIC_JPL/initialize_with_solar.py

@@ -5,8 +5,7 @@ import rasters as rt
 
 from rasters import Raster
 
-from .vegetation_conversion.vegetation_conversion import FVC_from_NDVI
-from .soil_heat_flux import calculate_SEBAL_soil_heat_flux
+from SEBAL_soil_heat_flux import calculate_SEBAL_soil_heat_flux
 
 from .constants import *
 from .soil_moisture_initialization import initialize_soil_moisture

STIC_JPL/iterate_with_solar.py

@@ -5,7 +5,7 @@ import rasters as rt
 
 from rasters import Raster
 
-from .soil_heat_flux import calculate_SEBAL_soil_heat_flux
+from SEBAL_soil_heat_flux import calculate_SEBAL_soil_heat_flux
 
 from .constants import *
 from .canopy_air_stream import calculate_canopy_air_stream_vapor_pressure