Abstract
This study analyzes the potential of deep learning using probabilistic artificial neural networks (ANNs) for postprocessing ensemble precipitation forecasts at four observation locations. We split the precipitation forecast problem into two tasks: estimating the probability of precipitation and predicting the hourly precipitation. We then compare the performance with classical statistical postprocessing (logistical regression and GLM). ANNs show a higher performance at three of the four stations for estimating the probability of precipitation and at all stations for predicting the hourly precipitation. Further, two more general ANN models are trained using the merged data from all four stations. These general ANNs exhibit an increase in performance compared to the station-specific ANNs at most stations. However, they show a significant decay in performance at one of the stations at estimating the hourly precipitation. The general models seem capable of learning meaningful interactions in the data and generalizing these to improve the performance at other sites, which also causes the loss of local information at one station. Thus, this study indicates the potential of deep learning in weather forecasting workflows.
Abstract
This study analyzes the potential of deep learning using probabilistic artificial neural networks (ANNs) for postprocessing ensemble precipitation forecasts at four observation locations. We split the precipitation forecast problem into two tasks: estimating the probability of precipitation and predicting the hourly precipitation. We then compare the performance with classical statistical postprocessing (logistical regression and GLM). ANNs show a higher performance at three of the four stations for estimating the probability of precipitation and at all stations for predicting the hourly precipitation. Further, two more general ANN models are trained using the merged data from all four stations. These general ANNs exhibit an increase in performance compared to the station-specific ANNs at most stations. However, they show a significant decay in performance at one of the stations at estimating the hourly precipitation. The general models seem capable of learning meaningful interactions in the data and generalizing these to improve the performance at other sites, which also causes the loss of local information at one station. Thus, this study indicates the potential of deep learning in weather forecasting workflows.
Abstract
Central Europe has experienced a sequence of unprecedented summer droughts since 2015, which had considerable effects on the functioning and productivity of natural and agricultural systems. Placing these recent extremes in a long-term context of natural climate variability is, however, constrained by the limited length of observational records. Here, we use tree-ring stable oxygen and carbon isotopes to develop annually resolved reconstructions of growing season temperature and summer moisture variability for central Europe during the past 2,000 years. Both records are independently interpolated across the southern Czech Republic and northeastern Austria to produce explicit estimates of the optimum agroclimatic zones, based on modern references of climatic forcing. Historical documentation of agricultural productivity and climate variability since 1090 CE provides strong quantitative verification of our new reconstructions. Our isotope records not only contain clear expressions of the Medieval (920-1000 CE) and Renaissance (early 16th century) droughts, but also the relative influence of temperature and moisture on hydroclimatic conditions during the first millennium (including previously reported pluvials during the early 3rd, 5th, and 7th centuries CE). We conclude, Czech agricultural production has experienced significant extremes over the past 2,000 years, which includes periods for which there are no modern analogues.
Abstract
Central Europe has experienced a sequence of unprecedented summer droughts since 2015, which had considerable effects on the functioning and productivity of natural and agricultural systems. Placing these recent extremes in a long-term context of natural climate variability is, however, constrained by the limited length of observational records. Here, we use tree-ring stable oxygen and carbon isotopes to develop annually resolved reconstructions of growing season temperature and summer moisture variability for central Europe during the past 2,000 years. Both records are independently interpolated across the southern Czech Republic and northeastern Austria to produce explicit estimates of the optimum agroclimatic zones, based on modern references of climatic forcing. Historical documentation of agricultural productivity and climate variability since 1090 CE provides strong quantitative verification of our new reconstructions. Our isotope records not only contain clear expressions of the Medieval (920-1000 CE) and Renaissance (early 16th century) droughts, but also the relative influence of temperature and moisture on hydroclimatic conditions during the first millennium (including previously reported pluvials during the early 3rd, 5th, and 7th centuries CE). We conclude, Czech agricultural production has experienced significant extremes over the past 2,000 years, which includes periods for which there are no modern analogues.
Abstract
While numerous collaborations exist between the atmospheric sciences research community and the U.S. National Weather Service (NWS), collaborative research field studies between undergraduate (UG) students at universities and the NWS are less common. The Summertime Canyon Observations and Research to Characterize Heat Extreme Regimes (SCORCHER) study was an UG student-driven research field campaign conducted in Palo Duro Canyon State Park, Texas, United States, during the summer of 2021. The SCORCHER campaign was mainly aimed at improving our basic scientific understanding of extreme heat, public safety, and forecasting applications, and creating an empowering UG educational field research experience. This “In Box” article highlights the collaborative study design, execution, and lessons learned.
Abstract
While numerous collaborations exist between the atmospheric sciences research community and the U.S. National Weather Service (NWS), collaborative research field studies between undergraduate (UG) students at universities and the NWS are less common. The Summertime Canyon Observations and Research to Characterize Heat Extreme Regimes (SCORCHER) study was an UG student-driven research field campaign conducted in Palo Duro Canyon State Park, Texas, United States, during the summer of 2021. The SCORCHER campaign was mainly aimed at improving our basic scientific understanding of extreme heat, public safety, and forecasting applications, and creating an empowering UG educational field research experience. This “In Box” article highlights the collaborative study design, execution, and lessons learned.
Abstract
Drought is a recurrent natural phenomenon, but there is concern that climate change may increase the frequency or severity of drought in Alaska. Because most common drought indices were designed for lower latitudes, it is unclear how effectively they characterize drought in Alaska’s diverse high-latitude climates. Here, we compare three commonly used meteorological drought indices (the standardized precipitation index (SPI), the standardized precipitation evapotranspiration index (SPEI), and the self-calibrating Palmer drought severity index (scPDSI) to each other and to streamflow across Alaska’s 13 climate divisions. All of the drought indices identify major droughts, but the severity of the drought varies depending on the index used. The SPI and the SPEI are more flexible and often better correlated with streamflow than the scPDSI, and we recommend using them. Although SPI and SPEI are very similar in energy-limited climates, the drought metrics do diverge in drier locations in recent years, and considering the impact of temperature on drought may grow more important in the coming decades. Hargreaves PET estimates appeared more physically realistic than the more commonly used Thornthwaite equation and are equally easy to calculate, so we suggest using the Hargreaves equation when PET is estimated from temperature. This study, one of the first to evaluate drought indices for high-latitude regions, has the potential to improve drought monitoring and representation within the United States Drought Monitor, leading to more informed decision-making during drought in Alaska, and it improves our ability to track changes in drought driven by rising temperatures.
Abstract
Drought is a recurrent natural phenomenon, but there is concern that climate change may increase the frequency or severity of drought in Alaska. Because most common drought indices were designed for lower latitudes, it is unclear how effectively they characterize drought in Alaska’s diverse high-latitude climates. Here, we compare three commonly used meteorological drought indices (the standardized precipitation index (SPI), the standardized precipitation evapotranspiration index (SPEI), and the self-calibrating Palmer drought severity index (scPDSI) to each other and to streamflow across Alaska’s 13 climate divisions. All of the drought indices identify major droughts, but the severity of the drought varies depending on the index used. The SPI and the SPEI are more flexible and often better correlated with streamflow than the scPDSI, and we recommend using them. Although SPI and SPEI are very similar in energy-limited climates, the drought metrics do diverge in drier locations in recent years, and considering the impact of temperature on drought may grow more important in the coming decades. Hargreaves PET estimates appeared more physically realistic than the more commonly used Thornthwaite equation and are equally easy to calculate, so we suggest using the Hargreaves equation when PET is estimated from temperature. This study, one of the first to evaluate drought indices for high-latitude regions, has the potential to improve drought monitoring and representation within the United States Drought Monitor, leading to more informed decision-making during drought in Alaska, and it improves our ability to track changes in drought driven by rising temperatures.
Abstract
Three levels of process-oriented model diagnostics are applied to evaluate the Global Ensemble Forecast System-version 12 (GEFSv12) reforecasts. The level-1 diagnostics are focused on model systematic errors, which reveals that precipitation onset over tropical oceans occurs too early in terms of column water vapor accumulation. Since precipitation acts to deplete water vapor, this results in prevailing negative biases of precipitable water in the tropics. It is also associated with over-transport of moisture into the mid-and upper-troposphere, leading to a dry bias in the lower troposphere and a wet bias in the mid-upper troposphere. The level-2 diagnostics evaluate some major predictability sources on the extended-range time scale: the Madden-Julian Oscillation (MJO) and North American weather regimes. It is found that the GEFSv12 can skillfully forecast the MJO up to 16 days ahead in terms of the Real-time Multivariate MJO indices (bivariate correlation ≥ 0.6) and can reasonably represent the MJO propagation across the Maritime Continent. The weakened and less coherent MJO signals with increasing forecast lead-times may be attributed to humidity biases over the Indo-Pacific warm pool region. It is also found that the weather regimes can be skillfully predicted up to 12 days ahead with persistence comparable to the observation. In the level-3 diagnostics, we examined some high-impact weather systems. The GEFSv12 shows reduced mean biases in tropical cyclone genesis distribution and improved performance in capturing tropical cyclone interannual variability, and mid-latitude blocking climatology in the GEFSv12 also shows a better agreement with the observations than in the GEFSv10.
Abstract
Three levels of process-oriented model diagnostics are applied to evaluate the Global Ensemble Forecast System-version 12 (GEFSv12) reforecasts. The level-1 diagnostics are focused on model systematic errors, which reveals that precipitation onset over tropical oceans occurs too early in terms of column water vapor accumulation. Since precipitation acts to deplete water vapor, this results in prevailing negative biases of precipitable water in the tropics. It is also associated with over-transport of moisture into the mid-and upper-troposphere, leading to a dry bias in the lower troposphere and a wet bias in the mid-upper troposphere. The level-2 diagnostics evaluate some major predictability sources on the extended-range time scale: the Madden-Julian Oscillation (MJO) and North American weather regimes. It is found that the GEFSv12 can skillfully forecast the MJO up to 16 days ahead in terms of the Real-time Multivariate MJO indices (bivariate correlation ≥ 0.6) and can reasonably represent the MJO propagation across the Maritime Continent. The weakened and less coherent MJO signals with increasing forecast lead-times may be attributed to humidity biases over the Indo-Pacific warm pool region. It is also found that the weather regimes can be skillfully predicted up to 12 days ahead with persistence comparable to the observation. In the level-3 diagnostics, we examined some high-impact weather systems. The GEFSv12 shows reduced mean biases in tropical cyclone genesis distribution and improved performance in capturing tropical cyclone interannual variability, and mid-latitude blocking climatology in the GEFSv12 also shows a better agreement with the observations than in the GEFSv10.
Abstract
IMERG provides the state-of-the-art satellite-based precipitation estimates that combine observations from multiple satellite platforms. This study evaluates IMERG products by examining hydrologic simulations of streamflow at a range of spatial scales. The main objective of this study is to assess the predictive utility of the near real-time product (IMERG-Early). The assessment also includes the IMERG-Final product that is not available in real time. The authors used MRMS precipitation estimates and USGS streamflow observation data as references for the precipitation and streamflow evaluations during a five-year period (2016–2020). The precipitation evaluation results show that IMERG-Early yields significant overestimations, particularly during warm months, with higher variability in its conditional distributions, whereas the performance of IMERG-Final seems unbiased. The authors performed hydrologic simulations using the Iowa Flood Center’s Hillslope Link Model with three precipitation forcing products i.e., MRMS, IMERG-Early, and IMERG Final. The simulation results reveal that IMERG-Early leads to high hit and false alarm rates due to its overestimation in precipitation and has almost no skill, as measured by the overall performance metric KGE, in streamflow prediction regarding basin scales ranging from 10 to 30,000 km2. This indicates that the product requires a bias correction before it is useful for real-time flood prediction. The streamflow prediction performance of IMERG-Final seems comparable to that of MRMS at spatial scales greater than 100 km2. This scale limitation is attributable to IMERG’s product spatial resolution that is inadequate to capture the small-scale variability of precipitation.
Abstract
IMERG provides the state-of-the-art satellite-based precipitation estimates that combine observations from multiple satellite platforms. This study evaluates IMERG products by examining hydrologic simulations of streamflow at a range of spatial scales. The main objective of this study is to assess the predictive utility of the near real-time product (IMERG-Early). The assessment also includes the IMERG-Final product that is not available in real time. The authors used MRMS precipitation estimates and USGS streamflow observation data as references for the precipitation and streamflow evaluations during a five-year period (2016–2020). The precipitation evaluation results show that IMERG-Early yields significant overestimations, particularly during warm months, with higher variability in its conditional distributions, whereas the performance of IMERG-Final seems unbiased. The authors performed hydrologic simulations using the Iowa Flood Center’s Hillslope Link Model with three precipitation forcing products i.e., MRMS, IMERG-Early, and IMERG Final. The simulation results reveal that IMERG-Early leads to high hit and false alarm rates due to its overestimation in precipitation and has almost no skill, as measured by the overall performance metric KGE, in streamflow prediction regarding basin scales ranging from 10 to 30,000 km2. This indicates that the product requires a bias correction before it is useful for real-time flood prediction. The streamflow prediction performance of IMERG-Final seems comparable to that of MRMS at spatial scales greater than 100 km2. This scale limitation is attributable to IMERG’s product spatial resolution that is inadequate to capture the small-scale variability of precipitation.
Abstract
Realistic ocean initial conditions are essential for coupled hurricane forecasts. This study focuses on the impact of assimilating high-resolution ocean observations for initialization of the Modular Ocean Model (MOM6) in a coupled configuration with the Hurricane Analysis and Forecast System (HAFS). Based on the Joint Effort for Data Assimilation Integration (JEDI) framework, numerical experiments were performed for the Hurricane Isaias (2020) case, a Category One hurricane, with use of underwater glider data sets and satellite observations. Assimilation of ocean glider data together with satellite observations provides opportunity to further advance understanding of ocean conditions and air-sea interactions in coupled model initialization and Hurricane forecast systems. This comprehensive data assimilation approach has led to a more accurate prediction of the salinity-induced barrier layer thickness that suppresses vertical mixing and sea surface temperature cooling during the storm. Increased barrier layer thickness enhances ocean enthalpy flux into the lower atmosphere and potentially increases tropical cyclone intensity. Assimilation of satellite observations demonstrates improvement in Hurricane Isaias’ intensity forecast. Assimilating glider observations with broad spatial and temporal coverage along Isaias’ track in addition to satellite observations further increase Isaias’ intensity forecast. Overall this case study demonstrates the importance of assimilating comprehensive marine observations to a more robust ocean and hurricane forecast under a unified JEDI-HAFS hurricane forecast system.
Abstract
Realistic ocean initial conditions are essential for coupled hurricane forecasts. This study focuses on the impact of assimilating high-resolution ocean observations for initialization of the Modular Ocean Model (MOM6) in a coupled configuration with the Hurricane Analysis and Forecast System (HAFS). Based on the Joint Effort for Data Assimilation Integration (JEDI) framework, numerical experiments were performed for the Hurricane Isaias (2020) case, a Category One hurricane, with use of underwater glider data sets and satellite observations. Assimilation of ocean glider data together with satellite observations provides opportunity to further advance understanding of ocean conditions and air-sea interactions in coupled model initialization and Hurricane forecast systems. This comprehensive data assimilation approach has led to a more accurate prediction of the salinity-induced barrier layer thickness that suppresses vertical mixing and sea surface temperature cooling during the storm. Increased barrier layer thickness enhances ocean enthalpy flux into the lower atmosphere and potentially increases tropical cyclone intensity. Assimilation of satellite observations demonstrates improvement in Hurricane Isaias’ intensity forecast. Assimilating glider observations with broad spatial and temporal coverage along Isaias’ track in addition to satellite observations further increase Isaias’ intensity forecast. Overall this case study demonstrates the importance of assimilating comprehensive marine observations to a more robust ocean and hurricane forecast under a unified JEDI-HAFS hurricane forecast system.
Abstract
This paper introduces a new tool for verifying tropical cyclone (TC) forecasts. Tropical cyclone forecasts made by operational centers and by numericalweather prediction (NWP) models have been objectively verified for decades. Typically, the mean absolute error (MAE) and/or MAE skill are calculated relative to values within the operations center’s best track. Yet, the MAE can be strongly influenced by outliers and yield misleading results. Thus, this paper introduces an assessment of consistency between the MAE skill as well as two other measures of forecast performance. This “consistency metric” objectively evaluates the forecast-error evolution as a function of lead time based on thresholds applied to the: 1) MAE skill, 2) median absolute error (MDAE) skill, and 3) the frequency of superior performance (FSP), which indicates how often one forecast outperforms another. The utility and applicability of the consistency metric is validated by applying it to four research and forecasting applications. Overall, this consistency metric is a helpful tool to guide analysis and increase confidence in results in a straightforward way. By augmenting the commonly-used MAE and MAE skill with this consistency metric and creating a single scorecard with consistency-metric results for TC track, intensity, and significant-windradii forecasts, the impact of observing systems, new modeling systems, or model upgrades on TC-forecast performance can be evaluated both holistically and succinctly. This could in turn help forecasters learn from challenging cases and accelerate and optimize developments and upgrades in NWP models.
Abstract
This paper introduces a new tool for verifying tropical cyclone (TC) forecasts. Tropical cyclone forecasts made by operational centers and by numericalweather prediction (NWP) models have been objectively verified for decades. Typically, the mean absolute error (MAE) and/or MAE skill are calculated relative to values within the operations center’s best track. Yet, the MAE can be strongly influenced by outliers and yield misleading results. Thus, this paper introduces an assessment of consistency between the MAE skill as well as two other measures of forecast performance. This “consistency metric” objectively evaluates the forecast-error evolution as a function of lead time based on thresholds applied to the: 1) MAE skill, 2) median absolute error (MDAE) skill, and 3) the frequency of superior performance (FSP), which indicates how often one forecast outperforms another. The utility and applicability of the consistency metric is validated by applying it to four research and forecasting applications. Overall, this consistency metric is a helpful tool to guide analysis and increase confidence in results in a straightforward way. By augmenting the commonly-used MAE and MAE skill with this consistency metric and creating a single scorecard with consistency-metric results for TC track, intensity, and significant-windradii forecasts, the impact of observing systems, new modeling systems, or model upgrades on TC-forecast performance can be evaluated both holistically and succinctly. This could in turn help forecasters learn from challenging cases and accelerate and optimize developments and upgrades in NWP models.
