Drought Forecasting Research Articles

Drought Forecasting: Application of Ensemble and Advanced Machine Learning Approaches

Drought Forecasting: Application of Ensemble and Advanced Machine Learning Approaches

Quantitative Causality, Causality-Aided Discovery, and Causal Machine Learning

It has been said, arguably, that causality analysis should pave a promising way to interpretable deep learning and generalization. Incorporation of causality into artificial intelligence algorithms, however, is challenged with its vagueness, nonquantitativeness, computational inefficiency, etc. During the past 18 years, these challenges have been essentially resolved, with the establishment of a rigorous formalism of causality analysis initially motivated from atmospheric predictability. This not only opens a new field in the atmosphere-ocean science, namely, information flow, but also has led to scientific discoveries in other disciplines, such as quantum mechanics, neuroscience, financial economics, etc., through various applications. This note provides a brief review of the decade-long effort, including a list of major theoretical results, a sketch of the causal deep learning framework, and some representative real-world applications pertaining to this journal, such as those on the anthropogenic cause of global warming, the decadal prediction of El Niño Modoki, the forecasting of an extreme drought in China, among others.

Meteorological Drought Detection and Forecast Using Standardized Precipitation Index and Univariate Distribution Models: Case Study of Bamako, Mali

Meteorological Drought Detection and Forecast Using Standardized Precipitation Index and Univariate Distribution Models: Case Study of Bamako, Mali

Economic value and latent demand for agricultural drought forecast: Emerging market for weather and climate information in Central-Southern Nigeria

Provision of weather and climate services are expected to improve the capacity for rural households’ preparedness and response plans to weather shocks. With increase in public investments in developing and communicating weather information on local scale in Nigeria, uncertainty in timescales that meet farmers’ needs and economic value of the information is still poorly understood. It is now a policy concern on whether farmers’ preferences and demands might increase its uptake. This study analyzed the economic value, latent demand, and emerging market of weather and climate information in Central-Southern Nigeria. Farm-level cross-sectional data reveals that 76% of the respondents were willing to pay for improved weather information and early warnings in taking climate smart decisions. Within farmers who showed positive responses, 86% would pay for sub-seasonal to seasonal weather information while 38% would pay for medium and short range weather information respectively. The economic value of sub-seasonal to seasonal weather information was estimated at N1600 ($3.60) per year per capita with total aggregated value of N1.3 billion ($2.9 m) yearly for the derived savannah area. Predictive total market value of N17.43billion ($39 m) would be obtained from improved weather information in Nigeria. Simulated results of 5% increase in the uptake with better dissemination channel through mobile phones in addition to robust farmers’ oriented features will generate additional annual market value at N86m ($193,360) for service providers. Large farm size, good farm-income, mobile phone dissemination channels, and location-specific information were drivers of farmers’ uptake decisions of weather information in the dry savannah area. The huge emerging market for improved weather information should be developed into a public–private market to efficiently facilitate uptake and use in Nigeria.

Comparison of LSTM network, neural network and support vector regression coupled with wavelet decomposition for drought forecasting in the western area of the DPRK

Comparison of LSTM network, neural network and support vector regression coupled with wavelet decomposition for drought forecasting in the western area of the DPRK

Evaluation of Hybrid Wavelet Models for Regional Drought Forecasting

Drought forecasting is essential for risk management and preparedness of drought mitigation measures. The present study aims to evaluate the effectiveness of the proposed hybrid technique for regional drought forecasting. Multiple Linear Regression (MLR), Artificial Neural Network (ANN), and two wavelet techniques, namely, Discrete Wavelet Transform (DWT) and Wavelet Packet Transform (WPT), were evaluated in drought forecasting up to a lead time of six months. Standard error metrics were used to select optimal model parameters, such as number of inputs, number of hidden neurons, level of decomposition, and number of mother wavelets. Additionally, the performance of various mother wavelets, including the Haar wavelet (db1) and 19 Daubechies wavelets (db1 to db20), were evaluated. The results indicated that the ANN model produced better forecasts than the MLR model, whereas the hybrid models outperformed both ANN and MLR models, which failed to predict the SPI values for a lead time greater than two months. The performance of all the models was found to improve as the timescale increased from 3 to 12 months. However, all the models’ performances deteriorated as the lead time increased. The hybrid WPT-MLR was the best model for the study area. The findings indicated that a hybrid WPT-MLR model could be used for drought early warning systems in the study area.

A point-surface fusion method for vegetation water content retrievalconsidering optimization of GNSS sites and modeling elements

The normalized microwave reflectance index (NMRI) based on global navigation satellite system (GNSS) interferometric reflectometry has been proven to reflect the changes in vegetation water content (VWC) effectively, but it is limited to point data. A spatially continuous NMRI product can be obtained by fusing NMRI data (point) and moderate-resolution imaging spectroradiometer (MODIS) data (surface). However, the time resolution of the existing results is limited to 16 d, and the research on the selection of modeling elements is not deep enough. In this paper, a point-surface fusion method for VWC retrieval considering the optimization of GNSS sites and elements is proposed. This method is aimed at using MODIS multi-band to synthesize vegetation indices with 8 d spatial-temporal resolution and establishing the initial element set by combining meteorological elements, followed by optimizing the GNSS sites and modeling elements for best modeling scheme, finally constructing the point-surface fusion method for VWC retrieval, and comprehensively evaluating the performance of the method. The results indicate that optimizing GNSS sites and modeling elements are particularly critical to improving modeling accuracy. Compared with other vegetation elements, normalized difference vegetation index (NDVI), gross primary productivity, and leaf area index are essential elements that affect the modeling effect. Among them, NDVI is the critical element. The NMRI products with 8 d/500 m resolution obtained by GA-BPNN can better reflect the change of VWC. Furthermore, the spatial performance of NMRI products is consistent with the fire forecast products and is suitable for drought and fire forecasts.

Development of wavelet-based Kalman Online Sequential Extreme Learning Machine optimized with Boruta-Random Forest for drought index forecasting

Drought is a stochastic and recurring hydrological natural hazard that occurs due to a shortage of precipitation over a period of time. Drought forecasting in water resources systems has an important role in reducing devastating ecological and social impacts. However, due to the fluctuating nature of drought indicator time series, their forecasting on a short time scale without advanced pre-processing is extremely challenging, and little research has been presented in this field. In this study, a new complementary machine learning (ML) approach, Kalman filter regression-based Online Sequential Extreme Learning Machine and (KOSELM) coupled with the Boundary Corrected Maximal Overlap Discrete Wavelet Transform (BC-MODWT-KOSELM), was implemented for forecasting one-month/three-month ahead Standardized Precipitation Evapotranspiration Index (i.e., SPEI 12 and SPEI 24). Here, the Kalman filter regression was employed to optimize the hyper-parameters of the Online Sequential Extreme Learning Machine (OSELM). Precipitation and potential evapotranspiration data from Bandar Abbas (warm semi-humid climate) and Rasmar (humid climate) synoptic stations, Iran were used in the analysis for a period of (1987–2019). In order to validate the BC-MODWT-KOSELM model, it was compared with two other well-known ML approaches, classical Extreme Learning Machine (ELM) and Generalized Regression Neural Network (GRNN) in both standalone and BC-MODWT-based complementary frameworks (i.e., BC-MODWT-ELM and BC-MODWT-GRNN). First, the original time series of the benchmark drought index was decomposed into the wavelet and scaling coefficients based on three mother wavelets and different decomposition levels. Afterward, the most significant lags were extracted using the Boruta-random forest (B-RF) feature selection. The BC-MODWT-KOSELM was identified as the superior forecasting model for SPEI 12 (t ＋ 1) with R = 0.9511 and RMSE = 0.3318; for SPEI 12 (t ＋ 3) with R = 0.8171 and RMSE = 0.6168; for SPEI 24 (t ＋ 1) with R = 0.9710 and RMSE = 0.2156; for SPEI 24 (t ＋ 3) with R = 0.9014 and RMSE = 0.3937 in the Bandar Abbas station. Besides, the BC-MODWT-KOSELM outperformed the comparative counterpart model for SPEI 12 (t ＋ 1 with R = 0.9401 and RMSE = 0.3371; SPEI 12 (t ＋ 3) with R = 0.8266 and RMSE = 0.5723; for SPEI 24 (t ＋ 1) with R = 0.9640 and RMSE = 0.2974; for SPEI 24 (t ＋ 3) with R = 0.9063 and RMSE = 0.4874 in the Ramsar station.

Enhanced Prediction and Determination of Hydrological Drought at Ungauged River Intake Stations under Changing Climate

Droughts, which are expected to worsen under global climate change, have major impacts on human life and the natural environment. In this study, an analysis system was established for predicting and determining hydrological drought conditions at ungauged water stations and in watersheds connected to municipal river water intake facilities. The aim was to help prevent drought damage or minimize its effects based on an immediate response to severe drought events. A system is presented for the selection of ungauged watersheds that take in river water, and three methodologies are proposed for identifying and forecasting hydrological drought conditions. Two South Korean pilot sites among the numerous ungauged water intake plants that lack local data collection facilities were selected as study areas. In addition, a roadmap for the establishment of standards for the determination of drought conditions in ungauged river basins was proposed. The methodologies introduced in this study assume nationwide expansion and construction. Their utilization can facilitate effective drought responses, based on drought forecasting and restricted water supply criteria for each phase of water intake, at local (and other) waterworks.

Hydrological Drought Forecasting Using a Deep Transformer Model

Hydrological drought forecasting is essential for effective water resource management planning. Innovations in computer science and artificial intelligence (AI) have been incorporated into Earth science research domains to improve predictive performance for water resource planning and disaster management. Forecasting of future hydrological drought can assist with mitigation strategies for various stakeholders. This study uses the transformer deep learning model to forecast hydrological drought, with a benchmark comparison with the long short-term memory (LSTM) model. These models were applied to the Apalachicola River, Florida, with two gauging stations located at Chattahoochee and Blountstown. Daily stage-height data from the period 1928–2022 were collected from these two stations. The two deep learning models were used to predict stage data for five different time steps: 30, 60, 90, 120, and 180 days. A drought series was created from the forecasted values using a monthly fixed threshold of the 75th percentile (75Q). The transformer model outperformed the LSTM model for all of the timescales at both locations when considering the following averages: MSE=0.11, MAE=0.21, RSME=0.31, and R2=0.92 for the Chattahoochee station, and MSE=0.06, MAE=0.19, RSME=0.23, and R2=0.93 for the Blountstown station. The transformer model exhibited greater accuracy in generating the same drought series as the observed data after applying the 75Q threshold, with few exceptions. Considering the evaluation criteria, the transformer deep learning model accurately forecasts hydrological drought in the Apalachicola River, which could be helpful for drought planning and mitigation in this area of contested water resources, and likely has broad applicability elsewhere.

Triple Collocation of Ground-, Satellite- and Land Surface Model-Based Surface Soil Moisture Products in Oklahoma—Part I: Individual Product Assessment

Improvements in soil moisture observations and modeling play a vital role in drought, water resources, flooding, and landslide management and forecasting. However, the lack of multisensor products that integrate different spatial scales (i.e., from 1 m2 to 102 km2) is a pressing need in the management and forecasting chain. Up to date, surface soil moisture estimates could be obtained through three primary approaches: (1) in situ measurements and their interpolations, (2) remote sensing observations, and (3) land surface model (LSM) outputs. Each source of soil moisture has its own spatiotemporal resolution, strengths, and weaknesses. Therefore, their correct interpretation and application require an in-depth understanding of their accuracy and appropriateness. In this study, we explore the utility of the triple collocation (TC) method for an independent assessment of three soil moisture products to characterize their uncertainty structures and make recommendations toward a potential product merge. The state of Oklahoma is an ideal domain to test the hypotheses of this work because of the presence of marked west-to-east gradients in climate, vegetation, and soils. The three target soil moisture products include (1) the remotely sensed microwave soil moisture active passive (SMAP) L3_SM_P_E (9 km, daily), (2) the physically based LSM estimates from NLDAS_NOAH0125_H (1/8°, hourly; Noah), and (3) the Oklahoma Mesonet ground sensor network (point, 30 min). The product assessment was conducted from April 2015 to July 2019. The results indicate that, in general, Mesonet and Noah are the most reliable products, although their performance varies geographically and by land cover type, reflecting the main spatiotemporal characteristics and scope of each product. Specifically, Mesonet provides the best estimates of volumetric soil moisture with a mean Pearson correlation coefficient of 0.805, followed by Noah with 0.747. However, Noah represents the true soil moisture variation better than the interpolated Mesonet product on the mesoscale, with an averaged RMSE of 0.026 m3⁄m3. Over different land cover types, Mesonet had the best performance in shrub/scrub, herbaceous, hay/pasture, and cultivated crops with an average correlation coefficient of 0.79, while Noah achieved the best performance in evergreen, mixed, and deciduous forests, with an average correlation coefficient of 0.74. The period-integrated TC intercomparison results over nine climate divisions indicated that Noah outperformed in the central, northeast, and east-central regions. TC provides not only a new perspective for comparatively assessing multisource soil moisture products but also a basis for objective data merging to capitalize on the strengths of multisensor, multiplatform soil moisture products.

Evaluating the predictability of eight Atmospheric-Oceanic signals affecting Iran’s Droughts, employing intelligence based and stochastic methods

The large-scale atmospheric-oceanic phenomena are among the main effective factors in the droughts in the Middle East, especially in Iran. Since these effects are usually delayed, their relevant signals can be useful for predicting droughts. As a result, the provision of a precise prediction of these signals can be efficient in increasing the drought prediction prospect. The current study predicts 8 cases of the most effective oceanic signals on the droughts which have been investigated in Iran. To do so, the problem-solving method with the time series prediction approach is based on the two model types intelligence-based (including multilayer perceptron [MLP] and support vector machine [SVM]) and stochastic (including Autoregressive Integrated Moving Average [ARIMA]) has been used. The model's input for each index included the time lags of the same index itself, which was determined by the autocorrelation function. Based on the evaluation criteria, the results were indicative of the weak predictability of the North Atlantic Oscillation (NAO) and Arctic Oscillation (AO), while the Extreme Eastern Tropical Pacific sea surface temperature (Niño [1 + 2]), East Central Tropical Pacific sea surface temperature (Niño [3 + 4]), and Oceanic Niño Index (ONI) were predicted with very good accuracy, and there is a high overlap between their predictions and observations (95.9 % < R2 < 99.3 %). In the extreme events also, the rate of normalized forecasting error for Niño (1 + 2), Niño (3 + 4), and ONI were in the medium (20–30 %), good (10–20 %), and excellent (0–10 %) ranges, respectively. The comparison between the models also indicates a partial superiority of the ARIMA stochastic model over the SVM and MLP models. The overall results of the study are indicative of the applicability of the predictions of the three mentioned indices as the inputs to increase precipitation and drought forecasting prospects in Iran (as well as all regions affected by them); which have the research value for further studies in terms of drought forecasting.

Status and prospects for drought forecasting: opportunities in artificial intelligence and hybrid physical–statistical forecasting

Despite major improvements in weather and climate modelling and substantial increases in remotely sensed observations, drought prediction remains a major challenge. After a review of the existing methods, we discuss major research gaps and opportunities to improve drought prediction. We argue that current approaches are top-down, assuming that the process(es) and/or driver(s) are known—i.e. starting with a model and then imposing it on the observed events (reality). With the help of an experiment, we show that there are opportunities to develop bottom-up drought prediction models—i.e. starting from the reality (here, observed events) and searching for model(s) and driver(s) that work. Recent advances in artificial intelligence and machine learning provide significant opportunities for developing bottom-up drought forecasting models. Regardless of the type of drought forecasting model (e.g. machine learning, dynamical simulations, analogue based), we need to shift our attention to robustness of theories and outputs rather than event-based verification. A shift in our focus towards quantifying the stability of uncertainty in drought prediction models, rather than the goodness of fit or reproducing the past, could be the first step towards this goal. Finally, we highlight the advantages of hybrid dynamical and statistical models for improving current drought prediction models. This article is part of the Royal Society Science+ meeting issue ‘Drought risk in the Anthropocene’.

A Place-Based Approach to Drought Forecasting in South-Central Oklahoma.

To assist water managers in south-central Oklahoma prepare for future drought, reliable place-based drought forecasts are produced. Past-, present-, and future-forecasted climate indices (Multivariate ENSO Index, Pacific Decadal Oscillation index, and Atlantic Multidecadal Oscillation index) and past and present Palmer Drought Severity Index (PDSI) are employed as predictor variables to forecast PDSI using a multivariate regression technique. PDSI is forecasted 18months in advance with sufficient skill to provide water managers early warning of drought. Using a training data set obtained from the periodJanuary 1901 to November 2021, a second-order model equation that contains, without any restriction, all the predictors and their interaction terms is built to predict drought intensity. Significant predictors are selected through stepwise regression, with cross-validation producing the simplest restricted model that describes the data well. PDSI values are predicted using 1000 fitted restricted models produced from bootstrapping, then averaged monthly. The technique found the best-fitting model and estimated the model coefficients that minimized the sum of squared deviations between the fitted model and the predictor variables. The adjusted R-squared value of the restricted model is large enough to explain an adequately accurate model, and relatively low values of error measures point to good predictive ability of the model. Although the model slightly overestimates the PDSI forecast maxima and minima, it necessarily captures the timing of the periods of severe to exceptional drought.

OneMap Drought Monitoring Analysis Based on Statistical Models

As the effects of droughts on agriculture and industrial water availability intensify with climate change, developing suitable drought prevention and mitigation measures has become increasingly important. However, measuring drought conditions using different indices leads to disjointed drought management responses by ministries and agencies. Additionally, indices based on only one variable are insufficient to accurately assess drought conditions. Therefore, creating and adopting a OneMap drought index would be beneficial in the assessment of drought conditions and the implementation of appropriate measures. In this study, we used multivariate statistical modeling using Bayesian principal component analysis to develop a OneMap drought index that unifies existing measures of drought conditions, including meteorological, agricultural, and hydrological drought indices. After evaluating the accuracy of the corrected OneMap drought index based on the self-organizing migrating algorithm optimization technique, it was found that the applicability of the OneMap drought index and its ability to regenerate drought were excellent for ground and satellite data. Therefore, the authors recommend implementing step-by-step drought management action plans using the integrated index to generate drought forecasts and warnings, thus promoting concerted and effective responses of local governments and authorities.

Application of a Hybrid CEEMD-LSTM Model Based on the Standardized Precipitation Index for Drought Forecasting: The Case of the Xinjiang Uygur Autonomous Region, China

Accurate forecasting of droughts can effectively reduce the risk of drought. We propose a hybrid model based on complementary ensemble empirical mode decomposition (CEEMD) and long short-term memory (LSTM) to improve drought prediction accuracy. Taking the Xinjiang Uygur Autonomous Region as an example, the prediction accuracy of the LSTM and CEEMD-LSTM models for the standardized precipitation index (SPI) on multiple timescales was compared and analyzed. Multiple evaluation metrics were used in the comparison of the models, such as the Nash–Sutcliffe efficiency (NSE). The results show that (1) with increasing timescale, the prediction accuracy of the LSTM and CEEMD-LSTM models gradually improves, and both reach their highest accuracy at the 24-month timescale; (2) the CEEMD can effectively stabilize the time-series, and the prediction accuracy of the hybrid model is higher than that of the single model at each timescale; and (3) the NSE values for the hybrid CEEMD-LSTM model at SPI24 were 0.895, 0.930, 0.908, and 0.852 for Fuhai, Kuerle, Yutian, and Hami station, respectively. This indicates the applicability of the hybrid model in the forecasting of drought.

Evaluation of Remote Sensing-Based Evapotranspiration Datasets for Improving Hydrological Model Simulation in Humid Region of East China

Conventional calibration methods used in hydrological modelling are based on runoff observations at the basin outlet. However, calibration with only runoff often produces reasonable runoff but poor results for other hydrological variables. Multi-variable calibration with both runoff and remote sensing-based evapotranspiration (ET) is developed naturally, due to the importance of ET and its data availability. This study compares two main calibration schemes: (1) calibration with only runoff (Scheme I) and (2) multi-variable calibration with both runoff and remote sensing-based ET (Scheme II). ET data are obtained from three remote sensing-based ET datasets, namely Penman–Monteith–Leuning (PML), FLUXCOM, and the Global Land Evaporation Amsterdam Model (GLEAM). The aforementioned calibration schemes are applied to calibrate the parameters of the Distributed Hydrology Soil Vegetation Model (DHSVM) through ε-dominance non-dominated sorted genetic algorithm II (ε-NSGAII). The results show that all three ET datasets have good performance for areal ET in the study area. The DHSVM model calibrated based on Scheme I produces acceptable performance in runoff simulation (Kling–Gupta Efficiency, KGE = 0.87), but not for ET simulation (KGE &lt; 0.7). However, reasonable simulations can be achieved for both variables based on Scheme II. The KGE value of runoff simulation can reach 0.87(0.91), 0.72(0.85), and 0.75(0.86) in the calibration (validation) period based on Scheme II (PML), Scheme II (FLUXCOM), and Scheme II (GLEAM), respectively. Simultaneously, ET simulations are greatly improved both in the calibration and validation periods. Furthermore, incorporating ET data into all three Scheme II variants is able to improve the performance of extreme flow simulations (including extreme low flow and high flow). Based on the improvement of the three datasets in extreme flow simulations, PML can be utilized for multi-variable calibration in drought forecasting, and FLUXCOM and GLEAM are good choices for flood forecasting.

Interannual Variability of Winter Rainfall in Upper Myanmar

Upper Myanmar region, roughly located between 21° 00' N and 28° 30' N latitude and 92° 10' E and 101° 11' E longitude, is the place where the winter cold season contributes ~2% of the annual total rainfall. The rainfall associated with Western disturbances is small in quantum but veritably important for the cold season crops, maintaining the glaciers over the Putao region, hydropower generation for the whole country and hazard of Jade mining of the Upper Myanmar area. This study aims to find interannual variability and related ocean-atmospheric pattern link with Upper Myanmar cold season rainfall by using great-resolution reanalysis data (ERA5) during 1990-2020. Correlation analysis to test the validation of ERA5 gridded data with the observed data from 25 stations across Myanmar, showed a strong correlation value in the same period that enough reliable for best analysis results. An anomalous anticyclonic (cyclonic) circulation persists over the southern part of the Bay of Bengal and South China sea during wet (dry) years. Also, the warming over the Indian Ocean and the cooling over the Tibetan plateau region correspond to south-north transport of moisture, ensuing in positive rainfall anomalies over the study region during winter. The wide patches of strong negative (positive) correlation are found over the Pacific Ocean, the Atlantic Ocean, Mediterranean Sea (MED), Arabian Sea (ARS), and Red Sea (RED) during wet (dry) years. The link implies that NPO, SPO, and MED have an impact on the winter rainfall inter-annual variability. In addition, the cooling (warming) over the Indochina and western Pacific regions influences the Hadley and Walker circulation bringing above (below) normal rainfall, respectively, over Upper Myanmar. The reply of indices (PO, MED, NINO3.4, IOD, and WDs) on winter rainfall, necessary to further investigation. The complete analysis of winter rainfall aids in the understanding of past extreme events as well as the forecasting and monitoring of drought and floods in Upper Myanmar.

Solving transparency in drought forecasting using attention models.

Droughts are one of the most devastating and recurring natural disaster due to a multitude of reasons. Among the different drought studies, drought forecasting is one of the key aspects of effective drought management. The occurrence of droughts is related to a multitude of factors which is a combination of hydro-meteorological and climatic factors. These variables are non-linear in nature, and neural networks have been found to effectively forecast drought. However, classical neural nets often succumb to over-fitting due to various lag components among the variables and therefore, the emergence of new deep learning and explainable models can effectively solve this problem. The present study uses an Attention-based model to forecast meteorological droughts (Standard Precipitation Index) at short-term forecast range (1-3 months) for five sites situated in Eastern Australia. The main aim of the work is to interpret the model outcomes and examine how a deep neural network achieves the forecasting results. The plots show the importance of the variables along with its short-term and long-term dependencies at different lead times. The results indicate the importance of large-scale climatic indices at different sequence dependencies specific to the study site, thus providing an example of the necessity to build a spatio-temporal explainable AI model for drought forecasting. The use of such interpretable models would help the decision-makers and planners to use data-driven models as an effective measure to forecast droughts as they provide transparency and trust while using these models.

How Predictable is Short-Term Drought in the Northeastern United States?

Abstract We investigated the predictability (forecast skill) of short-term droughts using the Palmer drought severity index (PDSI). We incorporated a sophisticated data training (of decadal range) to evaluate the improvement of forecast skill of short-term droughts (3 months). We investigated whether the data training of the synthetic North American Multi-Model Ensemble (NMME) climate has some influence on enhancing short-term drought predictability. The central elements are the merged information among PDSI and NMME with two postprocessing techniques. 1) The bias correction–spatial disaggregation (BC-SD) method improves spatial resolution by using a refined soil information introduced in the available water capacity of the PDSI calculation to assess water deficit that better estimates drought variability. 2) The ensemble model output statistic (EMOS) approach includes systematically trained decadal information of the multimodel ensemble simulations. Skill of drought forecasting improves when using EMOS, but BC-SD does not increase the forecast skill when compared with an analysis using BC (low spatial resolution). This study suggests that predictability forecast of drought (PDSI) can be extended without any change in the core dynamics of the model but instead by using the sophisticated EMOS postprocessing technique. We pointed out that using NMME without any postprocessing is of limited use in the suite of model variations of the NMME, at least for the U.S. Northeast. From our analysis, 1 month is the most extended range we should expect, which is below the range of the seasonal scale presented with EMOS (2 months). Thus, we propose a new design of drought forecasts that explicitly includes the multimodel ensemble signal.