A Granger causality analysis of groundwater patterns over a half-century

Groundwater depletion in many areas of the world has been broadly attributed to irrigation. However, more formal, data-driven, causal mechanisms of long-term groundwater patterns have not been assessed. Here, we conducted the first Granger causality analysis to identify the “causes” of groundwater patterns using the rice-producing parishes of Louisiana, USA, as an example. Trend analysis showed a decline of up to 6 m in groundwater level over 51 years. We found that no single cause explained groundwater patterns for all parishes. Causal linkages were noted between groundwater and area harvested, number of irrigation wells, summer precipitation totals, and drought. Bi-directional linkages were noted between groundwater and rice yield, suggesting feedback between both time series. Causal linkages were absent between groundwater and many drivers where significant correlations were noted, highlighting the importance of using robust causal relationships over illusive correlations to detect the cause. These results advance our understanding of groundwater dynamics and can reveal a key connection between food and groundwater.