CYGNSS Fractional Inundation¶

The UCAR/CU CYGNSS inundation maps provide fractional surface water extent derived from NASA's Cyclone Global Navigation Satellite System (CYGNSS), a constellation of eight microsatellites launched in December 2016. Each satellite carries a GNSS-Reflectometry (GNSS-R) receiver that records L-band GPS signals reflected from the Earth's surface. Because the transmitters are already on orbit, GNSS-R receivers are inexpensive to build and fly, which makes constellations — and therefore short revisit times — practical in a way that single-satellite radars are not. Surface reflectivity (Γ) at L-band is sensitive to the dielectric constant and micro-scale roughness of the reflecting surface, both of which change sharply when a landscape floods.

Inundation fraction is retrieved by treating the observed reflectivity of a 3 km grid cell as the weighted average of the reflectivity of its inundated and un-inundated fractions, with the weights equal to the respective cell fractions. Inverting this relationship requires ancillary estimates of soil moisture (SMAP L3 Enhanced), soil surface roughness, water surface roughness, and an empirical correction for the attenuating effect of above-ground biomass. Soil and water roughness are parameterized as spatially varying but temporally static fields calibrated against 2018 observations; soil moisture varies through time.

Because the eight satellites sample the surface pseudo-randomly and cannot completely cover it within a few days, observations are aggregated over 3-day windows and spatially interpolated using the Previously-Observed Behavior Interpolation (POBI) method, which derives historical relationships between reflectivity in neighboring grid cells and applies them to unobserved cells. POBI is an exact interpolator — observed values are unchanged — so users who prefer to avoid interpolated data can restrict their analysis to observed cells using the interpolation flag band.

Data span March 2017 to present at 3 km resolution, with daily files representing the mean inundation fraction for the current day and previous two days. Coverage is limited to ±38 degrees latitude by the CYGNSS orbital inclination.

Available Datasets¶

Dataset Details¶

Band Descriptions¶

Image Properties¶

Uncertainty and Known Limitations¶

Data Access¶

The source netCDF files are hosted on the COSMIC data server. The Earth Engine collection hosts the same retrievals converted to Cloud-Optimized GeoTIFF and updated to present day.

Citation¶

Chew, C., Small, E., Huelsing, H. 2023. Flooding and inundation maps using interpolated CYGNSS reflectivity observations. Remote Sensing of Environment 293: 113598. https://doi.org/10.1016/j.rse.2023.113598

Related publications:

• Chew, C.; Small, E. 2020. Estimating inundation extent using CYGNSS data: A conceptual modeling study. Remote Sensing of Environment 246: 111869. https://doi.org/10.1016/j.rse.2020.111869
• Chew, C. 2021. Spatial interpolation based on previously-observed behavior: a framework for interpolating spaceborne GNSS-R data from CYGNSS. Journal of Spatial Science. https://doi.org/10.1080/14498596.2021.1942253
• Chew, C.; Small, E. 2020. Description of the UCAR/CU Soil Moisture Product. Remote Sensing 12(10): 1558. https://doi.org/10.3390/rs12101558
• Chew, C.; Reager, J.T.; Small, E. 2018. CYGNSS data map flood inundation during the 2017 Atlantic hurricane season. Scientific Reports 8. https://doi.org/10.1038/s41598-018-27673-x
• Chew, C.; Small, E. 2018. Soil moisture sensing using spaceborne GNSS reflections: comparison of CYGNSS reflectivity to SMAP soil moisture. Geophysical Research Letters 45: 4049-4057. https://doi.org/10.1029/2018GL077905

Earth Engine Snippet¶

var cygnss = ee.ImageCollection("projects/sat-io/open-datasets/CYGNSS/INUNDATION");

// Scale to fractional inundation and mask fill values
var toFraction = function(image) {
  return image.select(['inundation_1day', 'inundation_3day'])
              .updateMask(image.select('inundation_3day').neq(255))
              .divide(100)
              .copyProperties(image, ['system:time_start', 'doy', 'year']);
};

// Anomalies carry a +100 bias
var toAnomaly = function(image) {
  return image.select('inundation_anomaly_3day')
              .updateMask(image.select('inundation_anomaly_3day').neq(255))
              .subtract(100).divide(100)
              .copyProperties(image, ['system:time_start']);
};

// Restrict to directly observed cells only
var observedOnly = function(image) {
  return image.updateMask(image.select('interpolation_flag').eq(0));
};

var palette = ['f7f4e9', 'c8c88c', '7ab88a', '2b8cbe', '08306b'];
Map.addLayer(cygnss.filterDate('2024-08-01', '2024-09-01').map(toFraction).mean(),
             {min: 0, max: 1, bands: ['inundation_3day'], palette: palette},
             'Mean fractional inundation, Aug 2024');

Sample Code: https://code.earthengine.google.com/?scriptPath=users/sat-io/awesome-gee-catalog-examples:hydrology/CYGNSS-INUNDATION

License¶

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License (CC BY-SA 4.0).

Created by: Clara Chew, Eric Small, and Hannah Huelsing — University Corporation for Atmospheric Research (UCAR) and University of Colorado Boulder, with funding from the NASA Terrestrial Hydrology Program

Curated in GEE by: Samapriya Roy

Keywords: inundation, flooding, surface water, CYGNSS, GNSS-R, reflectometry, wetlands, hydrology

Last updated: 2026-08-05

Ask AI