No abstract

No abstract

Flood prediction across scales and more specifically in ungauged areas remains still a great challenge that limits the efficiency of flood risk mitigation strategies and disaster preparedness.Building upon the recent success of Machine Learning (ML) models on streamflow prediction, this work presents a prototype ML-based framework for flood warning and flood peak prediction. The fundamental elements of the proposed system consist of a) a LSTM model for classifying storm events to threat/no-threat given a threshold based on the 90th flow percentile and b) the flood peak prediction models. The selected ML-models for flood peak prediction are the Histogram based Gradient Boosting Regressor and the Random Forest. One of the strengths, and reason for selection, of these decision-tree models is their degree of interpretability. This is exploited in the study to help us spatially disentangle the role of both the static and dynamic drivers of flood peak response. Our analysis is presented for 18 distinct hydroclimatic regions across the contiguous US. Results reveal a significant regional dependence on both predictive performance and dominant flood predictors, which emphasize the variability in the complexity of a catchment's hydrologic behavior as well as its impact on modeling flood response. Evaluation of the drivers of flood peaks noted distinct dependencies among the dynamic and static predictors considered in our models for flood peaks of different severity. Specifically, low-moderate flood events showed a clear preponderance for the static catchment attributes over dynamic predictors like precipitation whereas precipitation was the dominant driver for the high severity flood peaks. The proposed flood peak prediction models were compared against a state-of-the-art LSTM model and were shown to consistently outperform in ungauged basins. Overall, the proposed system classified storms correctly for >85% in all cases and exhibited a percent relative difference in flood peak estimates of <30% in most cases.

No abstract

&lt;p&gt;Global hydrologic climate assessments posit increasing flood risk. Hydrologic forecasting is critical in both gauged and ungauged basins having implications not only for hazard assessments and the development of mitigation strategies but also for informing the design and operation of critical infrastructure. The hydrology community grapples with the need to predict floods particularly in ungauged basins where the absence of continuous and spatially representative precipitation and streamflow data are enunciated.&amp;#160;&lt;/p&gt;&lt;p&gt;Global precipitation observations from satellite constellations combined with recent advancements of hydrologic forecasting with machine-learning (ML) models, offer an attractive solution for addressing flood prediction in ungauged regions. Towards that end, in this work, we investigate a) the performance of ML flood prediction models integrated with satellite precipitation estimates and b) the transferability/applicability of ML models trained in data rich regions for flood prediction in ungauged regions. We use NASA IMERG precipitation dataset for ML-based predictions and we train the ML models for ~600 catchments from different hydroclimatic zones in Contiguous US. The performance of the ML-IMERG predictions are then evaluated for a large number of catchments (~1000) in the UK, Brazil, Chile and Australia. Predictive performance is evaluated with respect to climate and catchment characteristics in each region. Results suggest that despite the variability in the performance across regions, there is great promise on the integration of global satellite precipitation estimates with ML models for flood prediction in ungauged basins. &amp;#160;&lt;/p&gt;