High Resolution Soil Moisture

Authors

Noemi Vergopolan, Eliot Feibush

Project Summary

Soil moisture reveals a lot about a place, from which areas of a farm need irrigation to which neighborhoods are likely to flood. Satellites and in-ground sensors provide valuable data on soil moisture, but these methods can be expensive and scarce when it comes to getting detailed information about frequently- changing conditions. To complement these data sources, Noemi Vergopolan of Princeton University combined satellite and model simulations to provide high-resolution soil moisture estimates across the entire continental U.S.

These images show the distribution of moisture in the top 5 centimeters of soil during wet winter conditions (top) and dry summer conditions (bottom). With a temporal resolution of 3 hours and a spatial resolution of 30 meters, the methodology allows users to easily zoom in for an estimate of the conditions at particular places and times (inset boxes show a 33-kilometer area at native 30-meter resolution). The results were simulated using the HydroBlocks land surface model, which aggregates big data from various Earth science data sets into a format that can be efficiently accessed and used.

*Published in 2020 CASC Brochure*

The Coalition for Academic Scientific Computation (CASC) has over 90 member institutions, including Princeton University.  CASC provides a forum for best practice computing and data services to advance academic research. They recognize that visualization plays a key role in exploring, verifying, and communicating the large amount of data generated by computer simulations. 

CASC publishes an annual brochure of images produced by its members. This visualization was selected for the 2020 CASC brochure and can be found on page 13.

Technology/Software Used

A new workflow, originally developed for climate model data, was used to create the visualization. Written entirely in Python, the visualizer reads variables from netCDF files and color-codes each grid point.  The workflow is tailored to f(x,y,t) data.  Standardizing on the netCDF format enabled displaying data from the HydroBlocks model with no changes to the visualization software.