Search Results
You are looking at 1 - 4 of 4 items for
- Author or Editor: Manabendra Saharia x
- Refine by Access: All Content x
Abstract
In the hydrological sciences, the outstanding challenge of regional modeling requires to capture common and event-specific hydrologic behaviors driven by rainfall spatial variability and catchment physiography during floods. The overall objective of this study is to develop robust understanding and predictive capability of how rainfall spatial variability influences flood peak discharge relative to basin physiography. A machine-learning approach is used on a high-resolution dataset of rainfall and flooding events spanning 10 years, with rainfall events and basins of widely varying characteristics selected across the continental United States. It overcomes major limitations in prior studies that were based on limited observations or hydrological model simulations. This study explores first-order dependencies in the relationships between peak discharge, rainfall variability, and basin physiography, and it sheds light on these complex interactions using a multidimensional statistical modeling approach. Among different machine-learning techniques, XGBoost is used to determine the significant physiographical and rainfall characteristics that influence peak discharge through variable importance analysis. A parsimonious model with low bias and variance is created that can be deployed in the future for flash flood forecasting. The results confirm that, although the spatial organization of rainfall within a basin has a major influence on basin response, basin physiography is the primary driver of peak discharge. These findings have unprecedented spatial and temporal representativeness in terms of flood characterization across basins. An improved understanding of subbasin scale rainfall spatial variability will aid in robust flash flood characterization as well as with identifying basins that could most benefit from distributed hydrologic modeling.
Abstract
In the hydrological sciences, the outstanding challenge of regional modeling requires to capture common and event-specific hydrologic behaviors driven by rainfall spatial variability and catchment physiography during floods. The overall objective of this study is to develop robust understanding and predictive capability of how rainfall spatial variability influences flood peak discharge relative to basin physiography. A machine-learning approach is used on a high-resolution dataset of rainfall and flooding events spanning 10 years, with rainfall events and basins of widely varying characteristics selected across the continental United States. It overcomes major limitations in prior studies that were based on limited observations or hydrological model simulations. This study explores first-order dependencies in the relationships between peak discharge, rainfall variability, and basin physiography, and it sheds light on these complex interactions using a multidimensional statistical modeling approach. Among different machine-learning techniques, XGBoost is used to determine the significant physiographical and rainfall characteristics that influence peak discharge through variable importance analysis. A parsimonious model with low bias and variance is created that can be deployed in the future for flash flood forecasting. The results confirm that, although the spatial organization of rainfall within a basin has a major influence on basin response, basin physiography is the primary driver of peak discharge. These findings have unprecedented spatial and temporal representativeness in terms of flood characterization across basins. An improved understanding of subbasin scale rainfall spatial variability will aid in robust flash flood characterization as well as with identifying basins that could most benefit from distributed hydrologic modeling.
Abstract
Flash floods, a subset of floods, are a particularly damaging natural hazard worldwide because of their multidisciplinary nature, difficulty in forecasting, and fast onset that limits emergency responses. In this study, a new variable called “flashiness” is introduced as a measure of flood severity. This work utilizes a representative and long archive of flooding events spanning 78 years to map flash flood severity, as quantified by the flashiness variable. Flood severity is then modeled as a function of a large number of geomorphological and climatological variables, which is then used to extend and regionalize the flashiness variable from gauged basins to a high-resolution grid covering the conterminous United States. Six flash flood “hotspots” are identified and additional analysis is presented on the seasonality of flash flooding. The findings from this study are then compared to other related datasets in the United States, including National Weather Service storm reports and a historical flood fatalities database.
Abstract
Flash floods, a subset of floods, are a particularly damaging natural hazard worldwide because of their multidisciplinary nature, difficulty in forecasting, and fast onset that limits emergency responses. In this study, a new variable called “flashiness” is introduced as a measure of flood severity. This work utilizes a representative and long archive of flooding events spanning 78 years to map flash flood severity, as quantified by the flashiness variable. Flood severity is then modeled as a function of a large number of geomorphological and climatological variables, which is then used to extend and regionalize the flashiness variable from gauged basins to a high-resolution grid covering the conterminous United States. Six flash flood “hotspots” are identified and additional analysis is presented on the seasonality of flash flooding. The findings from this study are then compared to other related datasets in the United States, including National Weather Service storm reports and a historical flood fatalities database.
Abstract
Prediction, and thus preparedness, in advance of flood events is crucial for proactively reducing their impacts. In the summer of 2012, Beijing, China, experienced extreme rainfall and flooding that caused 79 fatalities and economic losses of $1.6 billion. Using rain gauge networks as a benchmark, this study investigated the detectability and predictability of the 2012 Beijing event via the Global Hydrological Prediction System (GHPS), forced by the NASA Tropical Rainfall Measuring Mission (TRMM) Multisatellite Precipitation Analysis at near–real time and by the deterministic and ensemble precipitation forecast products from the NOAA Global Forecast System (GFS) at several lead times. The results indicate that the disastrous flooding event was detectable by the satellite-based global precipitation observing system and predictable by the GHPS forced by the GFS 4 days in advance. However, the GFS demonstrated inconsistencies from run to run, limiting the confidence in predicting the extreme event. The GFS ensemble precipitation forecast products from NOAA for streamflow forecasts provided additional information useful for estimating the probability of the extreme event. Given the global availability of satellite-based precipitation in near–real time and GFS precipitation forecast products at varying lead times, this study demonstrates the opportunities and challenges that exist for an integrated application of GHPS. This system is particularly useful for the vast ungauged regions of the globe.
Abstract
Prediction, and thus preparedness, in advance of flood events is crucial for proactively reducing their impacts. In the summer of 2012, Beijing, China, experienced extreme rainfall and flooding that caused 79 fatalities and economic losses of $1.6 billion. Using rain gauge networks as a benchmark, this study investigated the detectability and predictability of the 2012 Beijing event via the Global Hydrological Prediction System (GHPS), forced by the NASA Tropical Rainfall Measuring Mission (TRMM) Multisatellite Precipitation Analysis at near–real time and by the deterministic and ensemble precipitation forecast products from the NOAA Global Forecast System (GFS) at several lead times. The results indicate that the disastrous flooding event was detectable by the satellite-based global precipitation observing system and predictable by the GHPS forced by the GFS 4 days in advance. However, the GFS demonstrated inconsistencies from run to run, limiting the confidence in predicting the extreme event. The GFS ensemble precipitation forecast products from NOAA for streamflow forecasts provided additional information useful for estimating the probability of the extreme event. Given the global availability of satellite-based precipitation in near–real time and GFS precipitation forecast products at varying lead times, this study demonstrates the opportunities and challenges that exist for an integrated application of GHPS. This system is particularly useful for the vast ungauged regions of the globe.
Abstract
To issue early warnings for the public to act, for emergency managers to take preventive actions, and for water managers to operate their systems cost-effectively, it is necessary to maximize the time horizon over which streamflow forecasts are skillful. In this work, we assess the value of medium-range ensemble precipitation forecasts generated with the Hydrologic Ensemble Forecast Service (HEFS) of the U.S. National Weather Service (NWS) in increasing the lead time and skill of streamflow forecasts for five headwater basins in the upper Trinity River basin in north-central Texas. The HEFS uses ensemble mean precipitation forecasts from the Global Ensemble Forecast System (GEFS) of the National Centers for Environment Prediction (NCEP). For comparative evaluation, we verify ensemble streamflow forecasts generated with the HEFS forced by the GEFS forecast with those forced by the short-range quantitative precipitation forecasts (QPFs) from the NWS West Gulf River Forecast Center (WGRFC) based on guidance from the NCEP’s Weather Prediction Center. We also assess the benefits of postprocessing the raw ensemble streamflow forecasts and evaluate the impact of selected parameters within the HEFS on forecast quality. The results show that the use of medium-range precipitation forecasts from the GEFS with the HEFS extends the time horizon for skillful forecasting of mean daily streamflow by 1–3 days for significant events when compared with using only the 72-h River Forecast Center (RFC) QPF with the HEFS. The HEFS forced by the GEFS also improves the skill of two-week-ahead biweekly streamflow forecast by about 20% over climatological forecast for the largest 1% of the observed biweekly flow.
Abstract
To issue early warnings for the public to act, for emergency managers to take preventive actions, and for water managers to operate their systems cost-effectively, it is necessary to maximize the time horizon over which streamflow forecasts are skillful. In this work, we assess the value of medium-range ensemble precipitation forecasts generated with the Hydrologic Ensemble Forecast Service (HEFS) of the U.S. National Weather Service (NWS) in increasing the lead time and skill of streamflow forecasts for five headwater basins in the upper Trinity River basin in north-central Texas. The HEFS uses ensemble mean precipitation forecasts from the Global Ensemble Forecast System (GEFS) of the National Centers for Environment Prediction (NCEP). For comparative evaluation, we verify ensemble streamflow forecasts generated with the HEFS forced by the GEFS forecast with those forced by the short-range quantitative precipitation forecasts (QPFs) from the NWS West Gulf River Forecast Center (WGRFC) based on guidance from the NCEP’s Weather Prediction Center. We also assess the benefits of postprocessing the raw ensemble streamflow forecasts and evaluate the impact of selected parameters within the HEFS on forecast quality. The results show that the use of medium-range precipitation forecasts from the GEFS with the HEFS extends the time horizon for skillful forecasting of mean daily streamflow by 1–3 days for significant events when compared with using only the 72-h River Forecast Center (RFC) QPF with the HEFS. The HEFS forced by the GEFS also improves the skill of two-week-ahead biweekly streamflow forecast by about 20% over climatological forecast for the largest 1% of the observed biweekly flow.
