Predicting flood stages in watersheds with different scales using hourly rainfall dataset: A high-volume rainfall features empowered machine learning approach.

Accurate prediction of instantaneous high lake water levels and flood flows (flood stages) from micro-catchments to big river basins are critical for flood forecasting. Lake Carl Blackwell, a small-watershed reservoir in the south-central USA, served as a primary case study due to its rich historical dataset. Bearing knowledge that both current and previous rainfall contributes to the reservoirs' water body, a series of hourly rainfall features were created to maximize predicting power, which include total rainfall amounts in the current hour, the past 2 h, 3 h, …, 600 h in addition to previous-day lake levels. Notedly, the rainfall features are the accumulated rainfall amounts from present to previous hours rather than the rainfall amount in any specific hour. Random Forest Regression (RFR) was used to score the features' importance and predict the flood stages along with Neural Network - Multi-layer Perceptron Regression (NN-MLP), Support Vector Regression (SVR), Extreme Gradient Boosting (XGBoost), and the ordinary multi-variant linear regression (MLR) together with dimension reduced linear models of Principal Component Regression (PCR) and Partial Least Square Regression (PLSR). The prediction accuracy for the lake flood stages can be as high as 0.95 in R2, 0.11 ft. in mean absolute error (MAE), and 0.21 ft. in root mean square error (RMSE) for the testing dataset by the RFR (NN-MLP performed equally well), with small accuracy decreases by the other two non-linear algorithms of XGBoost and SVR. The linear regressions with dimension reductions had the lowest accuracy. Furthermore, our approach demonstrated high accuracy and broad applicability for surface runoff and streamflow predictions across three different-sized watersheds from micro-catchment to big river basins in the region, with increases of predicting power from earlier rainfall for larger watersheds and vice versa.

Farm impacts of the 2019 Missouri River floods and economic valuation of flood risk reduction.

The 2019 Missouri River flood caused billions of dollars in damage to businesses, homes, and public infrastructure. Yet little is known about the farm-level effects of this event and farmers' perceptions of its causes. This study reports on the operational and financial setbacks farmers sustained because of the 2019 floods, as well as on their beliefs on the causes of these floods. It further explores farmers' willingness to pay (WTP) to avoid flood risks and the factors that condition it. The empirical application focuses on a sample of approximately 700 Missouri farmers operating near the Missouri River. Results show that yield loss, loss of growing crops, and inability to plant crops were the three most important consequences of flooding. Nearly 40% of the flood-affected farmers reported financial losses of $100,000 or more. Most respondents identified government decision makers as the cause of the 2019 floods, and many believe that government should prioritize flood control over other benefits (recreation and fish and wildlife habitat) the Missouri River system provides. The WTP results show that less than half of the surveyed farmers were willing to pay to avoid flood risks, with an average WTP estimated at $3 per $10,000 value of agricultural land. Subjective but not objective risk exposure influences WTP for flood risk reduction. Other important determinants of WTP are risk aversion, disutility from flood risks, and respondents' age, income, and education. Directions for policy to improve flood risk management in the Missouri River Basin are discussed.