Integrated geospatial Python libraries for efficient analysis of modern elevation measurements

NASA established the Surface Topography and Vegetation (STV) Incubation program to develop and mature the next-generation measurement approaches to precisely map Earth’s changing surface and overlying vegetation structure, and prepare for a dedicated satellite mission within the next decade. Over the past two decades, large archives of 3D surface elevation measurements by airborne and satellite instruments including LiDAR, altimeters, Synthetic Aperture Radar, and stereo optical imagery have been systematically collected, though not always in a coordinated way. Yet, many of these datasets are fortuitously acquired over the same location within a short temporal window (e.g., <1-14 days) and many are now publicly available and hosted on the cloud. In theory, this is a great opportunity to synthesize myriad elevation measurements for STV researchers, but in practice merging these datasets accurately for scientific analysis requires dealing with numerous data formats, complex 4D coordinate reference systems, and securing access to significant computational resources.

We are developing an open-source Python library to identify, curate, and efficiently process coincident elevation measurements spanning the last several decades. This work would not be possible without well-integrated geospatial libraries (e.g. Geopandas, Xarray, Dask), as well as emerging cloud-native data and metadata formats such as Cloud-Optimized Geotiff and STAC-GeoParquet. We will describe our work to-date and reflect on the process of collaborative development across libraries, on our increasing reliance on Cloud resources, and current and future research directions.