Manifold methods for assimilating geophysical and meteorological data in Earth system models and their components

摘要: Abstract A novel manifold method of reconstructing dynamically evolving spatial fields is presented for assimilating data from sensor networks in integrated land surface – subsurface, oceanic/lake models. The was developed based on the assumption that can be mapped onto an underlying differential manifold. In this study, proposed used to reconstruct meteorological forcing over Lake Michigan, bathymetry inland lake (Gull Lake), and precipitation Grand River watershed Michigan. first case hourly were reconstructed run a three-dimensional hydrodynamic model Michigan quantify improvement results use new method. second example, Gull measured scatter point using technique. refined improved bathymetry. last daily participation six-year period as input integrated, distributed hydrologic model. All three examples illustrate superior performance standard methods terms accuracy computational efficiency. Our also indicate cross-validation technique evaluate reconstruction lead misleading conclusions about their relative performance.