
This summer, I worked as a Graduate Fellow in the University of Washington’s Data Science for Social Good program. I decided to apply for this internship because of its focus on leveraging data science to address complex societal issues, its emphasis on ethical thinking, and the opportunity to engage with key stakeholders throughout the course of our project. Having previously worked in a role using data and research to inform policy, I was drawn to this position for its focus on an important climate-related topic.
Using Satellite Data to Understand Groundwater Changes in the Colorado River Basin
I worked on a team with three other graduate students, where our project focused on using satellite data to understand groundwater change in the Colorado River Basin (CRB). The CRB is facing an unprecedented water crisis due to climate change, ongoing drought, and unsustainable water practices. We had two goals for our project: (1) analyze groundwater change in the region; and (2) build a reproducible workflow that can be used to study groundwater change in other regions.
We chose to focus on groundwater–that is, freshwater that is within the saturated zones under Earth’s surface–because of its importance for human life, its significance in the hydrological cycle, and its absence in basin-wide policy and water governance. Historically, groundwater estimates have relied on in-situ measurements from well networks, a process that is both costly and difficult. While basin-wide policy for water allocation has focused on surface water allocation, groundwater management has largely been left to states. There are several reasons for this divergence, including historical lack of data about groundwater at the basin-level, state-variation in groundwater reliance and availability, and the complex interactions between groundwater and surface water.
Getting estimates for groundwater changes is no easy task. Our main data for this project came from NASA’s Gravity Recovery and Climate Experiment (GRACE) mission. GRACE is a collaboration between U.S. and German space agencies starting in 2002 that launched twin satellites into the Earth’s orbit, roughly 137 miles apart, that measures changes in the Earth’s gravity field. These measurements allow scientists to derive timely and precise estimates of the Earth’s total water storage at different points on its surface.

Source: Jet Propulsion Lab (accessed here)
In order to get estimates for groundwater changes, we combined GRACE data with pre-existing data on other freshwater sources – snow pack, soil moisture, and surface water such as lakes and rivers. By combining these with GRACE data, we could then obtain estimates of groundwater changes.
The Potential for these Data in Policy
Throughout the course of our project, we were able to speak with regional experts who gave us incredible insight into the context of groundwater and water management in the CRB. Our conversations gave us much of the policy and political context and the historical and legal backgrounds that we would not have had otherwise.
Water allocation dates back to the 1922 Colorado River Compact, an inter-state agreement among the seven U.S. basin states. This Compact divided the CRB into upper and lower basins and set specific annual amounts for each state, with upper basin states agreeing not to deplete the flow reaching the lower basin states and Mexico. Subsequent negotiations and agreements addressing water rights of Native American tribes and Mexico, who were left out of the 1922 Compact, have been ongoing to ensure participation in water management.

Source: Jet Propulsion Lab (accessed here)
The CRB has been facing drought and water scarcity issues for the last two decades. Key drivers of these trends are human-induced factors and natural occurrences, which have seen increased frequency and intensity due to climate change. First, the region has experienced prolonged periods of drought leading to reduced precipitation and water availability. Moreover, climate change has led to reduced snowpack, earlier snowmelt, and increased evaporation rates. Some climate scientists view this crisis as the beginning of aridification in the region. Increased demand in the region and agricultural demands have put substantial strain on the river. Arizona, for example, has historically relied on groundwater for parts of its development. Groundwater shortages have recently led to the state limiting construction around Phoenix. Finally, agriculture uses nearly 80% of the region’s water supply in a typical year, with livestock feed accounting for a majority of agricultural water usage.
In 2007, the Colorado River Basin Guidelines were established in response to ongoing drought conditions and the potential for water shortages in the CRB. More specifically, they were established to address the declining surface water levels in Lake Mead and Lake Powell. These guidelines are set to expire in 2026, providing an opportunity for states and stakeholders to renegotiate and update the agreements. Our hope is that our work can help provide awareness of these data and an easily implementable workflow to use these data to help inform research and policy decisions such as this one.
Conclusion
My experience in the program gave me new insight into a specific climate-related issue, its political and policy context, and the potential for new technology and data to be leveraged in research and policy. Though I was expecting the data analysis to be technically advanced, I was not expecting just how complicated it would be to work with and understand GRACE and other satellite data. Still, I am convinced of the potential for these data in policy and research despite their complicated nature. Finally, this experience highlighted for me the intersection of climate change and social issues, and I hope to be able to gain more experience studying and working on these topics.
Thank you to project leads Akshay Mehra and Sameer Shah, community engagement lead, Linda Méndez Barrientos, Data Scientist Lead Vaughn Iverson, eScience staff and support, and team members Aanchal Setia, Maia Powell, and Violet Huang for the opportunity to collaborate on this project and for your support.