WATERMELON

Satellite-based volumetric soil moisture estimation fusing microwave and optical remote sensing technologies.

Get to know about it
Try it out
Provide feedback

wide-area moderate meter-level gnss-ir and multispectral satellite data fusion for soil moisture estimation

Surface soil moisture from
Earth Observation data. Accompanied with NDVI and LST, providing comprehensive insights of soil moisture, vegetation health and surface temperature.

Accurate & cost-effective.

No infrastructure deployment.

Scalable to large areas.

In a nutshell watermelon seed 🍉

Research-based

Solution based on extensive research with real data and thoroughly studied monitoring metrics, exploiting the physical relationship between soil moisture and its signature in EO data and multiple climate variables.

WATERMELON current underlying technology is more advanced than its proof of concept. Some featured research that helped the initial proof of concept is:

[1] Msc. Thesis N. Padrón: Soil Moisture Estimation with GNSS Interferometric Reflectometry and Multispectral Satellite Model.

[2] N. Padrón and S.A. Vorobyov, “A GNSS-IR aided multispectral satellite data fusion for meter-level wide-area volumetric soil moisture estimation,” 50th IEEE Int. Conf. Acoustics, Speech, and Signal Processing, IEEE ICASSP’25, Hyderabad, India, Apr. 6–11, 2025, accepted. DOI: https://doi.org/10.1109/ICASSP49660.2025.10890009

Accuracy and Resolution

  • Moderate meter-level resolution with 10 and 30 meter/pixel.
  • Top root zone (surface) soil moisture estimates.
  • Evaluated with reference soil moisture dataset.
  • Accurate GNSS-IR soil moisture retrievals combined with moderate meter-level multispectral EO data for trustable, wide-area estimates.
  • Currently targeted to agriculture applications, aiding soil health and irrigation planning.
GNSS-IR: GPS| Galileo
GNSS Interferometric Reflectometry (GNSS-IR) is a passive microwave remote sensing technique capable of estimating climate variables from GNSS receivers.
Multispectral: Landsat | Copernicus
Multispectral satellite imagery covers various wavelengths of the electromagnetic spectrum, allowing to gather information about environmental features beyond the visible light.

Features

  • Landsat 8 & Sentinel 2
  • CSV download of long-term monthly averages.
  • Process up to 5 sites in parallel.
  • VSM accompanied with NDVI and LST to provide a more thorough result interpretation.
    • VSM: Volumetric Soil Moisture, reported in cubic cm of water per cubic cm of land.
      NDVI: Normalized Differential Vegetation Index, dimensionless.
      LST: Land Surface Temperature, reported in Celsius.
  • Clickable value per pixel, get accurate estimates in specific areas.
  • Dual panel, compare Volumetric Soil Moisture (VSM) together with NDVI and LST, or even contrast VSM in multiple times of the year.
  • Process up to 100 square km areas.
  • Estimation exploiting the benefits of microwave and multispectral technologies: GNSS-IR and Landsat/Sentinel.

How to use?

  • Explore the predefined cases in the Use Cases tab and click Run.
  • Watch the video for a brief demo for drawing a region and analyzing temporal and spatial soil moisture.
  • Have a look on the tabs and functionality description below.
  • Main Tabs
  • Use Case Selection
  • Draw Area
  • Multi-Site
  • Output
  • Save Results

Give a try to the prototype version!

Provide feedback

Your feedback would be really helpful!

As a prototype, research and improvement are still ongoing.
The following 5-question anonymous survey would help the possible development of future phases of this prototype, as well as gaining a broader insight about the applications of accurate soil moisture estimation in your field.



    YesNo



    DailyWeeklyMonthly



    NoYes, does not use but would benefit form itYes, already uses it





    Let’s connect on