O2.6.5 Add Hyperspectral Radiative Transfer to Canopy Model

[Espeer5]

Purpose

Our current radiative transfer model divides incoming radiation into 2 bands: the photosynthetically active band, and the near-infrared. band. This is a practice typical of older canopy models, but newer models implement much more granular radiative transfer models, banded into small wavelength ranges. Newer models also incorporate leaf optical properties to achieve more sophisticated modeling. The science for these models is included in the earlier CliMA Land Model.

This is part of ongoing work on the Big Leaf canopy model.

Cost/benefits/risks

The main cost of this effort is developer hours. Implementing this model will be nontrivial and will require time.

Another cost could be efficiency. Moving from 2 bands to many wavelength bands will increase the complexity of the radiative transfer model. This could, especially if poorly designed, cause the overall canopy model speed to decrease.

These costs are clearly outweighed by the increased scientific value of the more modern radiative transfer model.

Producers

@Espeer5 @braghiere

Components

The main deliverable is a new radiative transfer model option which can be plugged into the Canopy model, similar to the Beer model or the TwoStream model. This will need to interact with the photosynthesis and other Canopy components as TwoStream/Beer RT models do.

Inputs

The model will take as input maps of leaf optical properties and will need to take in the soil albedo. The model will ingest the downwelling radiation across the wavelength spectrum over the course of the simulation.

Results and deliverables

All science from the radiative transfer model of the earlier CliMA Land Model is implemented and available in ClimaLand.jl. This is used with all appropriate global input data in the Land model global runs, with realistic (though uncalibrated) results and no NANs over the globe.

Task breakdown

• Edit existing radiative transfer code to ingest wavelength spectrum rather than PAR/NIR from drivers

Reviewers

@braghiere @kmdeck @AlexisRenchon