# Remote estimation of geologic composition using interferometric synthetic-aperture radar in California's Central Valley

**Authors:** Kyongsik Yun, Kyra Adams, John Reager, Zhen Liu, Caitlyn Chavez, Michael Turmon, Thomas Lu  
**arXiv:** [2212.04813](https://arxiv.org/abs/2212.04813)  
**Date:** 2022-12-04  
**Summary page:** https://www.summarizepaper.com/en/paper/2212.04813v1/

## Key points

- California's Central Valley produces a quarter of the nation's food
- Excessive groundwater pumping has led to rapid land subsidence, with some areas sinking at rates as high as 20 cm per year
- Land subsidence has significant implications for infrastructure resilience and groundwater sustainability
- The authors aimed to identify specific regions in the Central Valley with different temporal dynamics of land displacement and determine their underlying geological composition
- They utilized interferometric synthetic aperture radar (InSAR)-based land deformation temporal changes and machine learning techniques to remotely estimate geologic composition
- Certain areas in the Central Valley exhibited potentially reversible land deformation due to elastic land compaction caused by coarser grain geologic compositions, such as Helm
- Random forest and deep neural network regression models established a significant correlation between InSAR-based land deformation and geologic composition
- Notable accuracy was achieved with just 1/4 sparse sampling to reduce spatial correlations among data points, indicating that the model could be generalized for indirect estimation of geologic composition in other regions.
- It is possible to estimate geologic composition using InSAR-based land deformation data remotely, offering an alternative to expensive and time-consuming in-situ measurements that may not be practical in some areas.
- This study provides valuable insights into how remote sensing technology can help address critical challenges facing California's Central Valley agriculture industry while improving infrastructure resilience and groundwater sustainability.

## Summary

California's Central Valley is a vital agricultural center that produces a quarter of the nation's food. However, excessive groundwater pumping has led to rapid land subsidence, with some areas sinking at rates as high as 20 cm per year. This phenomenon has significant implications for infrastructure resilience and groundwater sustainability. In this study, the authors aimed to identify specific regions in the Central Valley with different temporal dynamics of land displacement and determine their underlying geological composition. They then sought to remotely estimate geologic composition using interferometric synthetic aperture radar (InSAR)-based land deformation temporal changes and machine learning techniques. The authors discovered that certain areas in the Central Valley, such as Helm, exhibited potentially reversible land deformation due to elastic land compaction caused by coarser grain geologic compositions. By utilizing random forest and deep neural network regression models, they established a significant correlation between InSAR-based land deformation and geologic composition. The authors also achieved notable accuracy with just 1/4 sparse sampling to reduce spatial correlations among data points, indicating that their model could be generalized for indirect estimation of geologic composition in other regions. The results suggest that it is possible to estimate geologic composition using InSAR-based land deformation data remotely. This approach offers an alternative to expensive and time-consuming in-situ measurements of geologic composition that may not be practical in some areas. Furthermore, the generalizability of the model sheds light on high spatial resolution geologic composition estimation utilizing existing measurements. Overall, this study provides valuable insights into how remote sensing technology can help address critical challenges facing California's Central Valley agriculture industry while improving infrastructure resilience and groundwater sustainability.