Clustering and inundation maps based on CYGNSS reflectivity

March 16, 2023

Contact: [e-mail] (Pierre Zeiger)

These files present statistical parameters based on CYGNSS reflectivity derived
from the CYGNSS Level 1 Science Data Records, v3.0, which were further used to
produce a clustering map and an estimation of the fractional inundation.
The spatial resolution is 0.1° and the temporal resolution is 7 days. The
estimations cover the inter-tropical band from -37.5° to 37.5° latitude during a
1-year study period, from August 1, 2018 to July 31, 2019.

Here are the reference papers for this work:
1) Pierre Zeiger, Frédéric Frappart, José Darrozes, Catherine Prigent and Carlos
Jiménez. Analysis of CYGNSS coherent reflectivity over land for the
characterization of pan-tropical inundation dynamics. Remote Sensing of
Environment, 2022, https://doi.org/10.1016/j.rse.2022.113278

2) Pierre Zeiger, Frédéric Frappart, José Darrozes, Catherine Prigent, Carlos
Jiménez and Luc Bourrel. Weekly mapping of surface water extent in the
pan-tropical wetlands using spaceborne GNSS Reflectometry (submitted).

Two files are distributed:

a) LEGOS_ReflectivityCYGNSS_0.1deg_7D_2018-2019_v1.0.nc contains the statistical
parameters (weighted mean and standard deviation, median and median absolute
deviation, maximum, 75th and 90th percentile, number of observations) based
on CYGNSS reflectivity, gridded at 0.1° spatial resolution and 7-day temporal
resolution for the entire spatiotemporal coverage of the study. It should be
useful as a basis to understand the reflectivity signal in regions with varying
geophysical properties.

Dimensions: lon, lat, time
Variables: means (mean reflectivity), stds (std reflectivity), counts (number of
observations), maxs (maximum of reflectivity), medians (median reflectivity), mad
(median absolute deviation of reflectivity), per75 (75th percentile of reflectivity),
per90 (90th percentile of reflectivity)

b) LEGOS_WaterFractionCYGNSS_0.1deg_7D_2018-2019_v1.0.nc contains the more advanced
results, i.e. the clustering map permitting to delineate flooded regions and the
dynamic estimation of the fraction of water in 0.1° pixels. It should be useful to
compare with other flooding products and to support hydrological researches.

Dimensions: lon, lat, time
Variables: water_fraction (fractional inundation in the pixel), clustering_map
(output of the DTW clustering), pixel_size (size of 0.1° pxiels in km²)

This dataset is referenced, please use the dedicated citations.