A computational framework for hyperspectral radiative transfer modeling and deep learning emulation for global water quality applications at scale
Jeremy Kravitz, NASA Ames Research Center, jeremy.kravitz@nasa.gov (Presenter)
Liane Guild, NASA Ames Research Center, liane.s.guild@nasa.gov
Lisl Lain, CSIR, elain@csir.co.za
Stephen Mauceri, NASA JPL, steffen.mauceri@jpl.nasa.gov
Nick LaHaye, NASA JPL, nicholas.j.lahaye@jpl.nasa.gov
Laurel Hopkins, Oregon State University, hopkins.laurel@gmail.com
Ian Brosnan, NASA Ames Research Center, ian.g.brosnan@nasa.gov
Resulting socio-economic impacts of reduced water quality on lives and livelihoods are particularly felt in under-developed regions, with political instability an emerging consequence of water stress. Advertent monitoring and accurate forecasting is of critical importance towards appropriate and pragmatic management of coastal and inland aquatic resources. Thus, we present the spectral water inversion processor and emulator (SWIPE), a holistic computational framework for advanced forward and inverse optical modeling to enable accurate and rapid Earth Observation of aquatic ecosystems. SWIPE enables production of state-of-the-art hyperspectral synthetic radiometric data paired with full optical descriptions of the aquatic and atmospheric column relevant to global ocean, coastal, and inland aquatic ecosystems. A distributed equivalent algal populations (DEAP) model employs a two-layer (coated sphere) scattering code to derive the full suite of optical properties for a variety of inorganic and organic particles from first principles. Ranges of in-water and atmospheric constituent variability are selected to represent those of natural ecosystems globally, and corresponding hyperspectral reflectances are simulated via radiative transfer modeling in the visible and near-infrared spectrum. The synthetic dataset is leveraged to train deep learning models which can emulate the complex physical interactions underlying the nonlinear processes in Earth systems and directly produce a variety of biogeophysical properties of the water column with per-pixel uncertainty. Preliminary results show SWIPE retrieval algorithms able to decouple backscatter signals from inorganic and organic constituents which is expected to improve estimates of phytoplankton carbon stocks of coastal and inland systems and potentially close some gaps in global climate models. Our approach is computationally efficient, sensor agnostic, generalizes to various water types found globally, and provides quantified uncertainties. The synthetic dataset, bio-optical models, and inversion algorithms are planned to be open source and expected to facilitate mission design, science discovery, scaling efforts, and maximizing retrievable information content from various radiometric sensors.
Poster: Poster_Kravitz_1-50_175_35.pdf
Associated Project(s):
Poster Location ID: 1-50
Presentation Type: Poster
Session: Poster Session 1
Session Date: Tue (May 9) 5:00-7:00 PM
CCE Program: OBB