← Back to all interns
Lina presenting her poster

Adelina Rodriguez

Wesleyan University

Research Mentor: Nick Bond & Karin Bumbaco

Project: Comparative Analysis of NASA GRACE-FO L4 and SMAP-L4 Soil Moisture Products in Washington

Soil moisture is becoming an increasingly important parameter in drought indication. However, collecting soil moisture data on a large scale is not easy, due to the multitude of soil properties and environmental factors affecting water accumulation and drainage in soils. Especially here in Washington, where the diverse landscapes and soil types present an interesting challenge when measuring soil moisture, conducting data analysis, and reaching accurate conclusions. Additionally, there are numerous soil moisture products that collect data and analyze current conditions differently.

Figure 1: Example of strong SMAP performance in the Rainy Pass forested area during a dry summer (left). Example of strong GRACE performance in the same forested area during a wet spring (right).

Contradictory information from these products can lead to false interpretations of moisture levels and drought across the state. This summer, I investigated the performances of two gridded, satellite-derived soil moisture products from NASA (called GRACE and SMAP) in Washington. The goal of my project was to see if one product is more effective and accurate than the other at measuring soil moisture, and thus more useful for drought analysis. 

In order to analyze the performance of satellite products, I correlated the satellite data to in situ data measured at weather stations in Washington for four case studies, including spring 2021 and 2022 (April to June), summer 2018 (July to September), and fall 2016 (September to November). I also explored how these products represented contrasting landscapes, including forested regions, croplands, grasslands, and snowy areas. A large chunk of my summer was spent pre-processing the data sets I compiled from various sources. I used MATLAB to merge data into 3D arrays and plot dozens of time series graphs. I also created box and whisker plots to display correlation results (see Figure 2). 

Despite discrepancies in the datasets and limited analyses, the box and whisker plots show a general trend that SMAP outperforms GRACE in croplands, grasslands, and forested regions throughout the year, with the exception of snowy

Figure 2: Pearson correlations between satellite-derived products and in situ observations. Correlations are between daily variations in in situ data and respective satellite-derived data retrieved at the same spatial coordinates, during a dry spring (left) and a wet fall (right).

areas (see Figure 1, Figure 2). GRACE provided inaccurate soil moisture data in most regions during all seasons. This is most likely due to the differences in data collection technology. The SMAP satellite uses radar and radiometers to sense microwaves and measure brightness levels. The GRACE

double satellites use accelerometers to measure gravitational anomalies in changing water masses. When considering the large scale of Washington, SMAP satellites are able to pick up smaller changes in soil moisture. 

The main takeaway from my project is that drought coordinators, agricultural interests, and public officials should take land type and seasonality into careful consideration and gather soil moisture information from multiple sources. There is still plenty of work to be done on making soil moisture monitoring more cohesive in Washington. 

Project Introduction

Research Poster

Lina's research poster