Browse
Abstract
Urban climate model evaluation often remains limited by a lack of trusted urban weather observations. The increasing density of personal weather sensors (PWSs) make them a potential rich source of data for urban climate studies that address the lack of representative urban weather observations. In our study, we demonstrate that carefully quality-checked PWS data not only improve urban climate models’ evaluation but can also serve for bias correcting their output prior to any urban climate impact studies. After simulating near-surface air temperatures over London and southeast England during the hot summer of 2018 with the Weather Research and Forecasting (WRF) Model and its building Effect parameterization with the building energy model (BEP–BEM) activated, we evaluated the modeled temperatures against 402 urban PWSs and showcased a heterogeneous spatial distribution of the model’s cool bias that was not captured using official weather stations only. This finding indicated a need for spatially explicit urban bias corrections of air temperatures, which we performed using an innovative method using machine learning to predict the models’ biases in each urban grid cell. This bias-correction technique is the first to consider that modeled urban temperatures follow a nonlinear spatially heterogeneous bias that is decorrelated from urban fraction. Our results showed that the bias correction was beneficial to bias correct daily minimum, daily mean, and daily maximum temperatures in the cities. We recommend that urban climate modelers further investigate the use of quality-checked PWSs for model evaluation and derive a framework for bias correction of urban climate simulations that can serve urban climate impact studies.
Significance Statement
Urban climate simulations are subject to spatially heterogeneous biases in urban air temperatures. Common validation methods using official weather stations do not suffice for detecting these biases. Using a dense set of personal weather sensors in London, we detect these biases before proposing an innovative way to correct them with machine learning techniques. We argue that any urban climate impact study should use such a technique if possible and that urban climate scientists should continue investigating paths to improve our methods.
Abstract
Urban climate model evaluation often remains limited by a lack of trusted urban weather observations. The increasing density of personal weather sensors (PWSs) make them a potential rich source of data for urban climate studies that address the lack of representative urban weather observations. In our study, we demonstrate that carefully quality-checked PWS data not only improve urban climate models’ evaluation but can also serve for bias correcting their output prior to any urban climate impact studies. After simulating near-surface air temperatures over London and southeast England during the hot summer of 2018 with the Weather Research and Forecasting (WRF) Model and its building Effect parameterization with the building energy model (BEP–BEM) activated, we evaluated the modeled temperatures against 402 urban PWSs and showcased a heterogeneous spatial distribution of the model’s cool bias that was not captured using official weather stations only. This finding indicated a need for spatially explicit urban bias corrections of air temperatures, which we performed using an innovative method using machine learning to predict the models’ biases in each urban grid cell. This bias-correction technique is the first to consider that modeled urban temperatures follow a nonlinear spatially heterogeneous bias that is decorrelated from urban fraction. Our results showed that the bias correction was beneficial to bias correct daily minimum, daily mean, and daily maximum temperatures in the cities. We recommend that urban climate modelers further investigate the use of quality-checked PWSs for model evaluation and derive a framework for bias correction of urban climate simulations that can serve urban climate impact studies.
Significance Statement
Urban climate simulations are subject to spatially heterogeneous biases in urban air temperatures. Common validation methods using official weather stations do not suffice for detecting these biases. Using a dense set of personal weather sensors in London, we detect these biases before proposing an innovative way to correct them with machine learning techniques. We argue that any urban climate impact study should use such a technique if possible and that urban climate scientists should continue investigating paths to improve our methods.
Abstract
Daily expected wind power production from operating wind farms across North America are used to evaluate capacity factors (CF) computed using simulation output from the Weather Research and Forecasting (WRF) Model and to condition statistical models linking atmospheric conditions to electricity production. In Parts I and II of this work, we focus on making projections of annual energy production and the occurrence of electrical production drought. Here, we extend evaluation of the CF projections for sites in the Northeast, Midwest, southern Great Plains (SGP), and southwest U.S. coast (SWC) using statewide wind-generated electricity supply to the grid. We then quantify changes in the time scales of CF variability and the seasonality. Currently, wind-generated electricity is lowest in summer in each region except SWC, which causes a substantial mismatch with electricity demand. While electricity of residential heating may shift demand, research presented here suggests that summertime CF are likely to decline, potentially exacerbating the offset between seasonal peak power production and current load. The reduction in summertime CF is manifest for all regions except the SGP and appears to be linked to a reduction in synoptic-scale variability. Using fulfillment of 50% and 90% of annual energy production to quantify interannual variability, it is shown that wind power production exhibits higher (earlier fulfillment) or lower (later fulfillment) production for periods of over 10–30 years as a result of the action of internal climate modes.
Significance Statement
Electrical power system reassessment and redesign may be needed to aid efficient increased use of variable renewables in the generation of electricity. Currently wind-generated electricity in many regions of North America exhibits a minimum in summertime and hence is not well synchronized with electricity demand, which tends to be maximized in summer. Future projections indicate evidence of reductions in wind power during summer that would amplify this offset. However, electrification of heating may lead to increased wintertime demand, which would lead to greater synchronization.
Abstract
Daily expected wind power production from operating wind farms across North America are used to evaluate capacity factors (CF) computed using simulation output from the Weather Research and Forecasting (WRF) Model and to condition statistical models linking atmospheric conditions to electricity production. In Parts I and II of this work, we focus on making projections of annual energy production and the occurrence of electrical production drought. Here, we extend evaluation of the CF projections for sites in the Northeast, Midwest, southern Great Plains (SGP), and southwest U.S. coast (SWC) using statewide wind-generated electricity supply to the grid. We then quantify changes in the time scales of CF variability and the seasonality. Currently, wind-generated electricity is lowest in summer in each region except SWC, which causes a substantial mismatch with electricity demand. While electricity of residential heating may shift demand, research presented here suggests that summertime CF are likely to decline, potentially exacerbating the offset between seasonal peak power production and current load. The reduction in summertime CF is manifest for all regions except the SGP and appears to be linked to a reduction in synoptic-scale variability. Using fulfillment of 50% and 90% of annual energy production to quantify interannual variability, it is shown that wind power production exhibits higher (earlier fulfillment) or lower (later fulfillment) production for periods of over 10–30 years as a result of the action of internal climate modes.
Significance Statement
Electrical power system reassessment and redesign may be needed to aid efficient increased use of variable renewables in the generation of electricity. Currently wind-generated electricity in many regions of North America exhibits a minimum in summertime and hence is not well synchronized with electricity demand, which tends to be maximized in summer. Future projections indicate evidence of reductions in wind power during summer that would amplify this offset. However, electrification of heating may lead to increased wintertime demand, which would lead to greater synchronization.
Abstract
Extreme snow ablation can greatly impact regional hydrology, affecting streamflow, soil moisture, and groundwater supplies. Relatively little is known about the climatology of extreme ablation events in the eastern United States, and the causal atmospheric forcing mechanisms behind such events. Studying the Susquehanna River basin over a 50-yr period, here we evaluate the variability of extreme ablation and river discharge events in conjunction with a synoptic classification and global-scale teleconnection pattern analysis. Results indicate that an average of 4.2 extreme ablation events occurred within the basin per year, where some 88% of those events resulted in an increase in river discharge when evaluated at a 3-day lag. Both extreme ablation and extreme discharge events occurred most frequently during instances of southerly synoptic-scale flow, accounting for 35.7% and 35.8% of events, respectively. However, extreme ablation was also regularly observed during high pressure overhead and rain-on-snow synoptic weather types. The largest magnitude of snow ablation per extreme event occurred during occasions of rain-on-snow, where a basinwide, areal-weighted 5.7 cm of snow depth was lost, approximately 23% larger than the average extreme event. Interannually, southerly flow synoptic weather types were more frequent during winter seasons when the Arctic and North Atlantic Oscillations were positively phased. Approximately 30% of the variance in rain-on-snow weather type frequency was explained by the Pacific–North American pattern. Evaluating the pathway of physical forcing mechanisms from regional events up through global patterns allows for improved understanding of the processes resulting in extreme ablation and discharge across the Susquehanna basin.
Significance Statement
The purpose of this study is to better understand how certain weather patterns are related to extreme snowmelt and streamflow events and what causes those weather patterns to vary with time. This is valuable information for informing hazard preparation and resource management within the basin. We found that weather patterns with southerly winds were the most frequent patterns responsible for extreme melt and streamflow, and those patterns occurred more often when the Arctic and North Atlantic Oscillations were in their “positive” configuration. Future work should consider the potential for these patterns, and related impacts, to change over time.
Abstract
Extreme snow ablation can greatly impact regional hydrology, affecting streamflow, soil moisture, and groundwater supplies. Relatively little is known about the climatology of extreme ablation events in the eastern United States, and the causal atmospheric forcing mechanisms behind such events. Studying the Susquehanna River basin over a 50-yr period, here we evaluate the variability of extreme ablation and river discharge events in conjunction with a synoptic classification and global-scale teleconnection pattern analysis. Results indicate that an average of 4.2 extreme ablation events occurred within the basin per year, where some 88% of those events resulted in an increase in river discharge when evaluated at a 3-day lag. Both extreme ablation and extreme discharge events occurred most frequently during instances of southerly synoptic-scale flow, accounting for 35.7% and 35.8% of events, respectively. However, extreme ablation was also regularly observed during high pressure overhead and rain-on-snow synoptic weather types. The largest magnitude of snow ablation per extreme event occurred during occasions of rain-on-snow, where a basinwide, areal-weighted 5.7 cm of snow depth was lost, approximately 23% larger than the average extreme event. Interannually, southerly flow synoptic weather types were more frequent during winter seasons when the Arctic and North Atlantic Oscillations were positively phased. Approximately 30% of the variance in rain-on-snow weather type frequency was explained by the Pacific–North American pattern. Evaluating the pathway of physical forcing mechanisms from regional events up through global patterns allows for improved understanding of the processes resulting in extreme ablation and discharge across the Susquehanna basin.
Significance Statement
The purpose of this study is to better understand how certain weather patterns are related to extreme snowmelt and streamflow events and what causes those weather patterns to vary with time. This is valuable information for informing hazard preparation and resource management within the basin. We found that weather patterns with southerly winds were the most frequent patterns responsible for extreme melt and streamflow, and those patterns occurred more often when the Arctic and North Atlantic Oscillations were in their “positive” configuration. Future work should consider the potential for these patterns, and related impacts, to change over time.
Abstract
Wildfire risk is greatest during high winds after sustained periods of dry and hot conditions. This paper is a statistical extreme-event risk attribution study that aims to answer whether extreme wildfire seasons are more likely now than under past climate. This requires modeling temporal dependence at extreme levels. We propose the use of transformed-linear time series models, which are constructed similarly to traditional autoregressive–moving-average (ARMA) models while having a dependence structure that is tied to a widely used framework for extremes (regular variation). We fit the models to the extreme values of the seasonally adjusted fire weather index (FWI) time series to capture the dependence in the upper tail for past and present climate. We simulate 10 000 fire seasons from each fitted model and compare the proportion of simulated high-risk fire seasons to quantify the increase in risk. Our method suggests that the risk of experiencing an extreme wildfire season in Grand Lake, Colorado, under current climate has increased dramatically relative to the risk under the climate of the mid-twentieth century. Our method also finds some evidence of increased risk of extreme wildfire seasons in Quincy, California, but large uncertainties do not allow us to reject a null hypothesis of no change.
Abstract
Wildfire risk is greatest during high winds after sustained periods of dry and hot conditions. This paper is a statistical extreme-event risk attribution study that aims to answer whether extreme wildfire seasons are more likely now than under past climate. This requires modeling temporal dependence at extreme levels. We propose the use of transformed-linear time series models, which are constructed similarly to traditional autoregressive–moving-average (ARMA) models while having a dependence structure that is tied to a widely used framework for extremes (regular variation). We fit the models to the extreme values of the seasonally adjusted fire weather index (FWI) time series to capture the dependence in the upper tail for past and present climate. We simulate 10 000 fire seasons from each fitted model and compare the proportion of simulated high-risk fire seasons to quantify the increase in risk. Our method suggests that the risk of experiencing an extreme wildfire season in Grand Lake, Colorado, under current climate has increased dramatically relative to the risk under the climate of the mid-twentieth century. Our method also finds some evidence of increased risk of extreme wildfire seasons in Quincy, California, but large uncertainties do not allow us to reject a null hypothesis of no change.
Abstract
Recent voyages of the Australian R/V Investigator across the remote Southern Ocean have provided unprecedented observations of precipitation made with both an Ocean Rainfall and Ice-Phase Precipitation Measurement Network (OceanRAIN) maritime disdrometer and a dual-polarization C-band weather radar (OceanPOL). This present study employs these observations to evaluate the Global Precipitation Measurement (GPM) Integrated Multi-satellitE Retrievals for GPM (IMERG) and the fifth major global reanalysis produced by ECMWF (ERA5) precipitation products. Working at a resolution of 60 min and 0.25° (∼25 km), light rain and drizzle are most frequently observed across the region. The IMERG product overestimated precipitation intensity when evaluated against the OceanRAIN but captured the frequency of occurrence well. Looking at the synoptic/process scale, IMERG was found to be the least accurate (overestimated intensity) under warm-frontal and high-latitude cyclone conditions, where multilayer clouds were commonly present. Under postfrontal conditions, IMERG underestimated the precipitation frequency. In comparison, ERA5’s skill was more consistent across various synoptic conditions, except for high pressure conditions where the precipitation frequency (intensity) was highly overestimated (underestimated). Using the OceanPOL radar, an area-to-area analysis (fractional skill score) finds that ERA5 has greater skill than IMERG. There is little agreement in the phase classification between the OceanRAIN disdrometer, IMERG, and ERA5. The comparisons are complicated by the various assumptions for phase classification in the different datasets.
Significance Statement
Our best quantitative estimates of precipitation over the remote, pristine Southern Ocean (SO) continue to suffer from a high degree of uncertainty, with large differences present among satellite-based and reanalysis products. New instrumentation on the R/V Investigator, specifically a dual-polarization C-band weather radar (OceanPOL) and a maritime disdrometer (OceanRAIN), provide unprecedented high-quality observations of precipitation across the SO that will aid in improving precipitation estimates in this region. We use these observations to evaluate the IMERG and ERA5 precipitation products. We find that, in general, IMERG overestimated precipitation intensity, but captured the frequency of occurrence well. In comparison, ERA5 was found to overestimate the frequency of precipitation. Using the OceanPOL radar, an area-to-area analysis finds that ERA5 has greater skill than IMERG.
Abstract
Recent voyages of the Australian R/V Investigator across the remote Southern Ocean have provided unprecedented observations of precipitation made with both an Ocean Rainfall and Ice-Phase Precipitation Measurement Network (OceanRAIN) maritime disdrometer and a dual-polarization C-band weather radar (OceanPOL). This present study employs these observations to evaluate the Global Precipitation Measurement (GPM) Integrated Multi-satellitE Retrievals for GPM (IMERG) and the fifth major global reanalysis produced by ECMWF (ERA5) precipitation products. Working at a resolution of 60 min and 0.25° (∼25 km), light rain and drizzle are most frequently observed across the region. The IMERG product overestimated precipitation intensity when evaluated against the OceanRAIN but captured the frequency of occurrence well. Looking at the synoptic/process scale, IMERG was found to be the least accurate (overestimated intensity) under warm-frontal and high-latitude cyclone conditions, where multilayer clouds were commonly present. Under postfrontal conditions, IMERG underestimated the precipitation frequency. In comparison, ERA5’s skill was more consistent across various synoptic conditions, except for high pressure conditions where the precipitation frequency (intensity) was highly overestimated (underestimated). Using the OceanPOL radar, an area-to-area analysis (fractional skill score) finds that ERA5 has greater skill than IMERG. There is little agreement in the phase classification between the OceanRAIN disdrometer, IMERG, and ERA5. The comparisons are complicated by the various assumptions for phase classification in the different datasets.
Significance Statement
Our best quantitative estimates of precipitation over the remote, pristine Southern Ocean (SO) continue to suffer from a high degree of uncertainty, with large differences present among satellite-based and reanalysis products. New instrumentation on the R/V Investigator, specifically a dual-polarization C-band weather radar (OceanPOL) and a maritime disdrometer (OceanRAIN), provide unprecedented high-quality observations of precipitation across the SO that will aid in improving precipitation estimates in this region. We use these observations to evaluate the IMERG and ERA5 precipitation products. We find that, in general, IMERG overestimated precipitation intensity, but captured the frequency of occurrence well. In comparison, ERA5 was found to overestimate the frequency of precipitation. Using the OceanPOL radar, an area-to-area analysis finds that ERA5 has greater skill than IMERG.
Abstract
Populated urban areas along many coastal regions are vulnerable to landfalling tropical cyclones (TCs). To the detriment of surface parameterizations in mesoscale models, the complexities of turbulence at high TC wind speeds in urban canopies are presently poorly understood. Thus, this study explores the impacts of urban morphology on TC-strength winds and boundary layer turbulence in landfalling TCs. To better quantify how urban structures interact with TC winds, large-eddy simulations (LESs) are conducted with the Cloud Model 1 (CM1). This implementation of CM1 includes immersed boundary conditions (IBCs) to represent buildings and eddy recycling to maintain realistic turbulent flow perturbations. Within the IBCs, an idealized coastal city with varying scales is introduced. TC winds impinge perpendicularly to the urbanized coastline. Numerical experiments show that buildings generate distinct, intricate flow patterns that vary significantly as the city structure is varied. Urban IBCs produce much stronger turbulent kinetic energy than is produced by conventional surface parameterizations. Strong effective eddy viscosity due to resolved eddy mixing is displayed in the wake of buildings within the urban canopy, while deep and enhanced effective eddy viscosity is present downstream. Such effects are not seen in a comparison LES using a simple surface parameterization with high roughness values. Wind tunneling effects in streamwise canyons enhance pedestrian-level winds well beyond what is possible without buildings. In the arena of regional mesoscale modeling, this type of LES framework with IBCs can be used to improve parameters in surface and boundary layer schemes to more accurately represent the drag coefficient and the eddy viscosity in landfalling TC boundary layers.
Significance Statement
This is among the first large-eddy simulation model studies to examine the impacts of tropical cyclone–like winds around explicitly resolved buildings. This work is a step forward in bridging the gap between engineering studies that use computational fluid dynamics models or laboratory experiments for flow through cities and mesoscale model simulations of landfalling tropical cyclones that use surface parameterizations specialized for urban land use.
Abstract
Populated urban areas along many coastal regions are vulnerable to landfalling tropical cyclones (TCs). To the detriment of surface parameterizations in mesoscale models, the complexities of turbulence at high TC wind speeds in urban canopies are presently poorly understood. Thus, this study explores the impacts of urban morphology on TC-strength winds and boundary layer turbulence in landfalling TCs. To better quantify how urban structures interact with TC winds, large-eddy simulations (LESs) are conducted with the Cloud Model 1 (CM1). This implementation of CM1 includes immersed boundary conditions (IBCs) to represent buildings and eddy recycling to maintain realistic turbulent flow perturbations. Within the IBCs, an idealized coastal city with varying scales is introduced. TC winds impinge perpendicularly to the urbanized coastline. Numerical experiments show that buildings generate distinct, intricate flow patterns that vary significantly as the city structure is varied. Urban IBCs produce much stronger turbulent kinetic energy than is produced by conventional surface parameterizations. Strong effective eddy viscosity due to resolved eddy mixing is displayed in the wake of buildings within the urban canopy, while deep and enhanced effective eddy viscosity is present downstream. Such effects are not seen in a comparison LES using a simple surface parameterization with high roughness values. Wind tunneling effects in streamwise canyons enhance pedestrian-level winds well beyond what is possible without buildings. In the arena of regional mesoscale modeling, this type of LES framework with IBCs can be used to improve parameters in surface and boundary layer schemes to more accurately represent the drag coefficient and the eddy viscosity in landfalling TC boundary layers.
Significance Statement
This is among the first large-eddy simulation model studies to examine the impacts of tropical cyclone–like winds around explicitly resolved buildings. This work is a step forward in bridging the gap between engineering studies that use computational fluid dynamics models or laboratory experiments for flow through cities and mesoscale model simulations of landfalling tropical cyclones that use surface parameterizations specialized for urban land use.
Abstract
Accurate evaluation of the long-range dependence in hydroclimatic time series is important for understanding its inherent characteristics. However, the reliability of its evaluation may be questioned, since different methods may yield various outcomes. In this study, we evaluate the performances of seven widely used methods for estimating long-range dependence: absolute moment estimation, difference variance estimation, residuals variance estimation, rescaled range estimation, periodogram estimation, wavelet estimation (WLE), and discrete second derivative estimation (DSDE). We examine the influences of six major factors: data length, mean value, three nonstationary components (trend, jump, and periodicity), and one stationary component (short-range dependence). Results from the Monte-Carlo experiments show that WLE and DSDE have greater credibility than the other five methods. They also reveal that data length, as well as stationary and nonstationary components, have notable influences on the evaluation of long-range dependence. Following it, we use the WLE and DSDE methods to evaluate the long-range dependence of precipitation during 1961–2015 on Tibetan Plateau. The results indicate that the precipitation variability mirrors the long-range dependence of the Indian summer monsoon, but with obvious spatial difference. This result is consistent with the observations made by previous studies, further confirming the superiority of the WLE and DSDE methods. The outcomes from this study have important implications for modeling and prediction of hydroclimatic time series.
Abstract
Accurate evaluation of the long-range dependence in hydroclimatic time series is important for understanding its inherent characteristics. However, the reliability of its evaluation may be questioned, since different methods may yield various outcomes. In this study, we evaluate the performances of seven widely used methods for estimating long-range dependence: absolute moment estimation, difference variance estimation, residuals variance estimation, rescaled range estimation, periodogram estimation, wavelet estimation (WLE), and discrete second derivative estimation (DSDE). We examine the influences of six major factors: data length, mean value, three nonstationary components (trend, jump, and periodicity), and one stationary component (short-range dependence). Results from the Monte-Carlo experiments show that WLE and DSDE have greater credibility than the other five methods. They also reveal that data length, as well as stationary and nonstationary components, have notable influences on the evaluation of long-range dependence. Following it, we use the WLE and DSDE methods to evaluate the long-range dependence of precipitation during 1961–2015 on Tibetan Plateau. The results indicate that the precipitation variability mirrors the long-range dependence of the Indian summer monsoon, but with obvious spatial difference. This result is consistent with the observations made by previous studies, further confirming the superiority of the WLE and DSDE methods. The outcomes from this study have important implications for modeling and prediction of hydroclimatic time series.
Abstract
We use a simple risk model for U.S. hurricane wind and surge economic damage to investigate the impact of projected changes in the frequencies of hurricanes of different intensities due to climate change. For average annual damage we find that changes in the frequency of category 4 storms dominate. For distributions of annual damage we find that changes in the frequency of category 4 storms again dominate for all except the shortest return periods. Sensitivity tests show that accounting for landfall, uncertainties and correlations leads to increases in damage estimates. When we propagate the distributions of uncertain frequency changes to give a best estimate of the changes in damage, the changes are moderate. When we pick individual scenarios from within the distributions of frequency changes, we find a significant probability of much larger changes in damage. The inputs on which our study depends are highly uncertain, and our methods are approximate, leading to high levels of uncertainty in our results. Also, the damage changes we consider are only part of the total possible change in hurricane damage due to climate change. Total damage change estimates would also need to include changes due to other factors, including possible changes in genesis, tracks, size, forward-speed, sea-level rise, rainfall and exposure. Nevertheless, we believe that our results give important new insights into U.S. hurricane risk under climate change.
Abstract
We use a simple risk model for U.S. hurricane wind and surge economic damage to investigate the impact of projected changes in the frequencies of hurricanes of different intensities due to climate change. For average annual damage we find that changes in the frequency of category 4 storms dominate. For distributions of annual damage we find that changes in the frequency of category 4 storms again dominate for all except the shortest return periods. Sensitivity tests show that accounting for landfall, uncertainties and correlations leads to increases in damage estimates. When we propagate the distributions of uncertain frequency changes to give a best estimate of the changes in damage, the changes are moderate. When we pick individual scenarios from within the distributions of frequency changes, we find a significant probability of much larger changes in damage. The inputs on which our study depends are highly uncertain, and our methods are approximate, leading to high levels of uncertainty in our results. Also, the damage changes we consider are only part of the total possible change in hurricane damage due to climate change. Total damage change estimates would also need to include changes due to other factors, including possible changes in genesis, tracks, size, forward-speed, sea-level rise, rainfall and exposure. Nevertheless, we believe that our results give important new insights into U.S. hurricane risk under climate change.
Abstract
Predicting drought severity is essential for drought management and early warning systems. Although numerous physical model-based and data-driven methods have been put forward for drought prediction, their abilities are largely constrained by data requirements and modeling complexity. There remains a challenging task to efficiently predict categorial drought, especially for the U.S. Drought Monitor (USDM). Aiming at this issue, multiple Markov chains for USDM-based categorial drought prediction are successfully proposed and evaluated in this paper. In particular, this study concentrated on how the Markov order, step size, and training set length affected prediction accuracy (PA). According to experiments from 2000 to 2021, it was found that the 1-step and first-order Markov models had the best accuracy in predicting droughts up to 4 weeks ahead. The PA steadily dropped with increasing step size, and the average accuracy at monthly scale was 88%. In terms of seasonal variability, summer (July–August) had the lowest PA while winter had the highest (January–February). In comparison with the western region, the PA in the eastern United States is 25% higher. Moreover, the length of the training set had an obvious impact on the PA of the model. The PA in 1-step prediction was 87% and 78% under 20- and 5-yr training sets, respectively. The results of the study showed that Markov models predicted categorical drought with high accuracy in the short term and showed different performances on time and space scales. Proper use of Markov models would help disaster managers and policy makers to put mitigation policies and measures into practice.
Abstract
Predicting drought severity is essential for drought management and early warning systems. Although numerous physical model-based and data-driven methods have been put forward for drought prediction, their abilities are largely constrained by data requirements and modeling complexity. There remains a challenging task to efficiently predict categorial drought, especially for the U.S. Drought Monitor (USDM). Aiming at this issue, multiple Markov chains for USDM-based categorial drought prediction are successfully proposed and evaluated in this paper. In particular, this study concentrated on how the Markov order, step size, and training set length affected prediction accuracy (PA). According to experiments from 2000 to 2021, it was found that the 1-step and first-order Markov models had the best accuracy in predicting droughts up to 4 weeks ahead. The PA steadily dropped with increasing step size, and the average accuracy at monthly scale was 88%. In terms of seasonal variability, summer (July–August) had the lowest PA while winter had the highest (January–February). In comparison with the western region, the PA in the eastern United States is 25% higher. Moreover, the length of the training set had an obvious impact on the PA of the model. The PA in 1-step prediction was 87% and 78% under 20- and 5-yr training sets, respectively. The results of the study showed that Markov models predicted categorical drought with high accuracy in the short term and showed different performances on time and space scales. Proper use of Markov models would help disaster managers and policy makers to put mitigation policies and measures into practice.
