I-Corps: Geospatial Trend Detection for Hydro-power and Critical Infrastructure Design

The broader impact/commercial potential of this I-Corps project is the development of a trend detection software tool to predict climate change risks related to water. Climate change risks manifest primarily through water, causing extremes (droughts and floods) and systematic shifts in the water systems’ time series. With robust trend detection tools, it may be possible to factor in climate change impacts when calculating the annual rate of change of loss and net economic returns from building large and expensive infrastructure projects, like dams. The proposed technology may identify the changing mean conditions of hydro-climatic (or other) variables that have a year-to-year variability, but on average could be shifting away from the historical mean behavior, called non-stationarity of mean. Some examples of infrastructure that may benefit from this technology include engineering design estimations of hydro-climate extreme return periods (e.g., droughts and floods) for pipe networks for water supply or wastewater, dams (hydro-power generation), and climate-impact on critical infrastructure (power grids, nuclear and mining sites). In addition, this technology may promote rapid adoption of climate resilient infrastructure design standards. This I-Corps project is based on the development of technology for robust trend detection, or “non-stationarity detection.” The proposed technology combines the results from several decades of independent literature in hydro-climate sciences and econometrics to achieve a step-change improvement in detecting non-stationary trends. In addition to geosciences, these developments are of fundamental importance in other applied science fields, including econometrics, biometrics, and psychometrics. Numerous variables in these domains are monitored for their changing behavior over time (trending behavior), but often their changes are also strongly correlated in time (predictably oscillation), which typically confounds the ability to detect a trending behavior. The proposed technology helps cut through this noise. In doing so, false negatives and false positives on non-stationarity detection were reduced by up to 30% and 400%, respectively. By using this information, operational and capital planning costs may be optimized by minimizing misplaced spending for climate resilient infrastructure, and maximizing spending where it is needed most. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.