WARNING: This code is in development, is being provided without support, and is subject to change at any time without notification

This repository provides Google Earth Engine Python API based implementation of the SSEBop ET model.

The Operational Simplified Surface Energy Balance (SSEBop) model computes daily total actual evapotranspiration (ETa) using land surface temperature (Ts), maximum air temperature (Ta) and reference ET (ETr or ETo). The SSEBop model does not solve all the energy balance terms explicitly; rather, it defines the limiting conditions based on "gray-sky" net radiation balance principles and an air temperature parameter. This approach predefines unique sets of "hot/dry" and "cold/wet" limiting values for each pixel, allowing an operational model setup and a relatively shorter compute time. More information on the GEE implementation of SSEBop is published in Senay2022 and Senay2023 with additional details and model assessment.

Basic SSEBop model implementation in Earth Engine:

The primary component of the SSEBop model is the Image() class. The Image class can be used to compute a single fraction of reference ET (ETf) image from a single input image. The Image class should generally be instantiated from an Earth Engine Landsat image using the collection specific methods listed below. ET image collections can be built by computing ET in a function that is mapped over a collection of input images. Please see the Example Notebooks for more details.

SSEBop ET can currently be computed for Landsat Collection 2 Level 2 (SR/ST) images from the following Earth Engine image collections:

• LANDSAT/LC09/C02/T1_L2
• LANDSAT/LC08/C02/T1_L2
• LANDSAT/LE07/C02/T1_L2
• LANDSAT/LT05/C02/T1_L2

Note: Users are encouraged to prioritize use of Collection 2 data where available. Collection 1 was produced by USGS until 2022-01-01, and maintained by Earth Engine until 2023-01-01. [More Information]

To instantiate the class for a Landsat Collection 2 SR/ST image, use the Image.from_landsat_c2_sr method.

The input Landsat image must have the following bands and properties:

The primary output of the SSEBop model is the fraction of reference ET (ETf). The actual ET (ETa) can then be computed by multiplying the Landsat-based ETf image with the reference ET (e.g. ETr from GRIDMET).

Example SSEBop ETa from Landsat:

import openet.ssebop as ssebop

landsat_img = ee.Image('LANDSAT/LC08/C02/T1_L2/LC08_044033_20170716')
et_fraction = ssebop.Image.from_landsat_c2_sr(landsat_img).et_fraction
et_reference = ee.Image('IDAHO_EPSCOR/GRIDMET/20170716').select('etr')
et_actual = et_fraction.multiply(et_reference)

SSEBop images can also be built manually by instantiating the class with an ee.Image with the following bands: 'lst' (land surface temperature [K]) and 'ndvi' (normalized difference vegetation index). The input image must have 'system:index' and 'system:time_start' properties (described above).

import openet.ssebop as ssebop

input_img = (
    ee.Image([ee.Image(lst), ee.Image(ndvi)])
    .rename(['lst', 'ndvi'])
    .set({
        'system:index': 'LC08_044033_20170716',
        'system:time_start': ee.Date.fromYMD(2017, 7, 16).millis()})
)
et_fraction = ssebop.Image(input_img).et_fraction

Detailed Jupyter Notebooks of the various approaches for calling the OpenET SSEBop model are provided in the "examples" folder.

• Computing daily ET for a single Landsat image
• Computing a daily ET image collection from Landsat image collection
• Computing monthly ET from a collection

The daily maximum air temperature (Tmax) is essential for establishing the maximum ET limit (cold boundary) as explained in Senay2017. Support for source options includes CIMIS, GRIDMET, DAYMET, and other custom Image Collections. See the model Image class docstrings for more information.

Default Asset ID: projects/usgs-ssebop/tmax/daymet_v4_mean_1981_2010 (Daily median from 1981-2010)

Land Surface Temperature is currently calculated in the SSEBop approach two ways:

• Landsat Collection 2 Level-2 (ST band) images directly. More information can be found at: USGS Landsat Collection 2 Level-2 Science Products

The SSEBop ET model uses dT as a predefined temperature difference between Thot and Tcold for each pixel. In SSEBop formulation, hot and cold limits are defined on the same pixel; therefore, dT actually represents the vertical temperature difference between the surface temperature of a theoretical bare/dry condition of a given pixel and the air temperature at the canopy level of the same pixel as explained in Senay2018. The input dT is calculated under "gray-sky" conditions and assumed not to change from year to year, but is unique for each day and location.

Default Asset ID: projects/usgs-ssebop/dt/daymet_median_v6

In order to correspond the maximum air temperature with cold/wet limiting environmental conditions, the SSEBop model uses a temperature correction coefficient (c factor, sometimes labeled interchangeably as Tcorr) uniquely calculated for each Landsat scene. This temperature correction component is uniquely developed for SSEBop using a Forcing and Normalizing Operation (FANO) method featuring a linear relation between a normalized land surface temperature difference and NDVI difference using the dT parameter and a proportionality constant.

Note: Tcorr refers to the pixel-based ratio of LST_cold and Tmax while c factor is a statistical value that represents a region such as a 5-km grid size (or larger) value.

More information on parameter design and model improvements using the FANO method can be found in Senay2023. Additional SSEBop model algorithm theoretical basis documentation can be found here.

The 'FANO' parameter (default) can be implemented dynamically for each Landsat scene within the SSEBop Image object using the following Tcorr source:

model_obj = model.Image.from_landsat_c2_sr(
    ee.Image('LANDSAT/LC08/C02/T1_L2/LC08_044033_20170716'),
    tcorr_source='FANO',
    )

The FANO parameterization allows the establishment of the cold boundary condition regardless of vegetation cover density, improving the performance and operational implementation of the SSEBop ET model in sparsely vegetated landscapes, dynamic growing seasons, and varying locations around the world.

The OpenET SSEBop python module can be installed via pip:

pip install openet-ssebop

• earthengine-api
• openet-core

Each OpenET model is stored in the "openet" folder (namespace). The model can then be imported as a "dot" submodule of the main openet module.

import openet.ssebop as ssebop

Please see the CONTRIBUTING.rst.