Journal of Hydrologic Engineering

The joint behavior of drought characteristics under climate change is evaluated using the copula method, which has recently attained popularity in the analysis of complex hydrologic systems with correlated variables. Trivariate copulas are applied, in this study, to analyze the major drought variables, including duration, severity, and intensity, in Oregon’s Upper Klamath River Basin. Among the variables, results show that duration severity exhibits the strongest correlation, whereas duration intensity exhibits the least correlation. The impact of climate change on future droughts is evaluated using five general circulation models (GCMs) under one emission scenario. Despite more intense extreme events that are expected to occur in most parts of the globe in the future, the results of this study show that the Upper Klamath River Basin in the Pacific Northwest will experience less intense droughts affected by climate change. Compared with historical events, an overall decrease in drought duration and severity is estimated for this study area in the time period of 2020–2090 with maximum drought duration shown to decline from 8 to 5 months. Among the five GCMs employed in this study, GFDL-CM2.1 and CSIRO-MK3.0 are identified as the wettest and driest projections, respectively. High uncertainty associated with GCM products is demonstrated in the analysis of return period by means of bivariate copulas. However, all projections result in larger return periods (i.e., less frequent droughts) compared with historical droughts during the reference period.

General circulation models (GCMs) are routinely used to simulate future climatic conditions. However, rainfall outputs from GCMs are highly uncertain in preserving temporal correlations, frequencies, and intensity distributions, which limits their direct application for downscaling and hydrological modeling studies. To address these limitations, raw outputs of GCMs or regional climate models are often bias corrected using past observations. In this paper, a methodology is presented for using a nested bias-correction approach to predict the frequencies and occurrences of severe droughts and wet conditions across India for a 48-year period (2050–2099) centered at 2075. Specifically, monthly time series of rainfall from 17 GCMs are used to draw conclusions for extreme events. An increasing trend in the frequencies of droughts and wet events is observed. The northern part of India and coastal regions show maximum increase in the frequency of wet events. Drought events are expected to increase in the west central, peninsular, and central northeast regions of India.

This work analyzes the response of springs fed by karst/fractured limestone aquifers, which extensively outcrop in central Italy, to climatic variations. In the central Italy area, climatic variations are represented by a general decrease in annual and winter precipitation. It is shown how groundwater regime, the discharge of springs, and their response to climate change depend to a great extent on the geologic and structural setting of the system. Some mountain springs of central Italy are local systems representing the overflow of a deeper regional flow feeding larger base springs, often of poor quality. A dynamic groundwater divide separates the recharge areas of base springs from those of local springs: if, due to low recharge, the piezometric surface lowers, the watershed moves toward systems located at higher altitudes, reducing their recharge areas. Therefore, local springs connected to a base flow are more vulnerable to climate change than those which are not. The Bagnara and Lupa springs, taken as examples, have similar mean discharges (approximately $120\mathrm{L}/\mathrm{s}$) and recharge areas with comparable lithological, topographical, and climate characteristics. Nevertheless, only the discharge of Bagnara, which is connected to a regional flow, fell dramatically during recent prolonged drought periods (e.g., 2001–2003 and 2006–2007).

Drought duration and severity are two main properties for characterizing droughts. These drought properties are mutually correlated and may have different marginal distributions. A bivariate (or joint) distribution of drought duration and severity is therefore needed that is capable of accommodating their different marginal distributions. This study proposes a method, based on entropy theory, for constructing the bivariate distribution of drought duration and severity with different marginal distribution forms. By specifying constraints for drought duration and severity, the entropy-based bivariate distribution can be derived and then marginal distributions can be obtained accordingly. Monthly streamflow data from Brazos River at Waco, Texas, are employed to illustrate the application of the proposed method to model drought duration and severity for drought analysis. The copula method is also applied for comparison with the proposed entropy method.

The onset of climate and land-use change is forcing water managers to develop new techniques in response to the changing environment. This study uses established techniques to incorporate both projected climate change and projected land-use change into a hydrologic model of the Occoquan watershed, which encompasses an area of approximately $\mathrm{1,550}{\mathrm{km}}^2$ in northern Virginia. The techniques used develop a future projection of weather that recreates the historic time series, including the drought of record, as influenced by climate change, thereby facilitating integration into existing water-management practices. Incorporating land use and using multiple metrics for analysis of low flow (drought) allows for the determination of variations between the historic and future model flows. This study revealed a likelihood of increased low-flow volumes for the Occoquan watershed from both climate and land-use change, the majority of which were produced from land-use change in combination with expanded reclaimed water supply. Also, the increases from climate change, although influencing measurable changes in flow patterns, were much less than those from land-use change.

Droughts produce a complex set of negative economic, environmental, and social impacts from regional to national scales. The drought impact can be quantified using drought index time series. This study uses the monthly standardized precipitation index (SPI) time series as a drought index. Drought characteristics, namely, drought duration, severity, interval time and minimum SPI values, were determined. Five hundred years of daily rainfall data were simulated for evaluating drought characteristics. Appropriate distributions were selected for modeling drought duration, interval time, drought severity, and minimum SPI value in different drought states. The drought episodes were quantified using multivariate copula methods. Several copulas from the Archimedean and metaelliptical families were applied to construct four-dimensional joint distributions. The dependence structure in each drought state was investigated and drought probabilities and return period were calculated and analyzed based on a four-dimensional copula using the upper Han River basin, China, as a study area.

This paper examines past drought and assesses future drought scenarios for the Arkansas Red River Basin using two common drought indexes, the Standardized Precipitation Index (SPI) and Palmer Drought Severity Index (PDSI). Historical climate data within the 1900–2009 time frame were used to derive the past drought index estimates. The projected climate data under two greenhouse gas emission scenarios from 16 global climate models (GCMs) after bias correction and statistical downscaling were applied in drought occurrence frequency and affected area projection. The results derived from the SPI and PDSI show that widespread droughts mainly took place in the 1910s, 1930s, 1950s, and 1960s in the Arkansas Red River Basin, which agrees well with the historical climate record. Both the SPI and PDSI project that more frequent and severe droughts will appear in the second part of the 21st century under both of the emissions scenarios. Future PDSI projects that more severe droughts will occur in the western parts of this basin under one scenario.

The identification and prediction of drought events depend on the integrity of the dataset employed. Streamflow is a good indicator of surface water availability and has dominated the literature on frequency analysis of hydrological droughts and water management. However, gauged measurements are impaired by climate and land use changes, especially in large, modified watersheds. Hence, their use in drought prediction is limited because they may violate the assumption of stationarity unless a naturalized observation series is obtained. In this paper, a land surface model is used to generate runoff in the Rio Grande/Río Bravo del Norte basin. Land use land cover is kept constant so that changes are a result of climatological variations. The river threads across several climatic zones; therefore the basin is divided into different regions for analysis. Using statistical theory of runs, water deficit duration and severity and drought interarrival time are extracted from 3-month Standardized Runoff Index series for each region and for the whole basin. Copulas are used to develop joint distribution functions of water deficit duration and severity. Nine copulas, two from the extreme value family and seven from the one-parameter families of the Archimedean copulas, are tested. Different copulas are deemed suitable for each region because they are subject to different climatologic conditions, which affect the nature of deficit events, especially at the extremes. Conditional probability models of duration and severity are developed and compared for two regions—one experiencing alpine climate at the headwaters and one in the semiarid portion of the basin. Further, conditional probabilities derived from model runoff are compared with that based on observed precipitation. Finally, univariate and conditional return periods based on duration and severity are computed for the basin and their significance in long-term water resources planning is discussed.

Droughts are characterized by drought indexes that measure the departures of meteorological and hydrological variables, such as precipitation and streamflow, from their long-term averages. Although many drought indexes have been proposed in the literature, most use predefined thresholds for identifying drought classes, ignoring the inherent uncertainties in characterizing droughts. This study employs a hidden Markov model (HMM) for the probabilistic classification of drought states. Apart from explicitly accounting for the time dependence in the drought states, the HMM-based drought index (HMM-DI) provides model uncertainty in drought classification. The proposed HMM-DI is used to assess drought characteristics in Indiana by using monthly precipitation and streamflow data. The HMM-DI results were compared to those from standard indexes and the differences in classification results from the two models were examined. In addition to providing the probabilistic classification of drought states, the HMM is suited for analyzing the spatio-temporal characterization of droughts of different severities.

The El Niño Southern Oscillation (ENSO) climate-variability phenomenon greatly affects water availability in the Southeast United States. For example, it is well known that La Niña conditions bring drought to this region. In the past decade, several severe droughts have adversely impacted the water resources of many communities in this region, especially those that rely on surface-water systems. Because small- to mid-size communities are most vulnerable to climate variability, this study was undertaken to develop a climate variability–based community water deficit index (CWDI) for use by water managers in these communities. Although currently available drought indices can be useful tools for monitoring and forecasting purposes, they are not suitable for use in water-supply systems for small- to mid-size communities. The CWDI was conceptualized keeping in mind that it should (1) forecast hydrologic drought, (2) operate at a high spatial resolution, and (3) address both water supply and demand during droughts. The system dynamics–modeling software Structured Thinking Experiential Learning Laboratory with Animation was used to develop the modeling framework to estimate CWDI by evaluating differences in a community water supply and demand, and thus help forecast the severity of an impending drought. Another important feature of the CWDI is its ability to evaluate how drought-management policies can affect the severity of drought. The CWDI was tested in two small- to mid-size communities of this region (Auburn, Alabama, and Griffin, Georgia). The results indicate that the index not only can monitor drought in the studied water-supply systems, but can also forecast ENSO-induced hydrologic droughts in the region and can be used in drought planning.

This study borrows the measures developed for the operation of water resources systems as a means of characterizing droughts in a given region. It is argued that the common approach of assessing drought using a univariate measure (severity or reliability) is inadequate as decision makers need assessment of the other facets considered here. It is proposed that the joint distribution of reliability, resilience, and vulnerability (referred to as RRV in a reservoir operation context), assessed using soil moisture data over the study region, be used to characterize droughts. Use is made of copulas to quantify the joint distribution between these variables. As reliability and resilience vary in a nonlinear but almost deterministic way, the joint probability distribution of only resilience and vulnerability is modeled. Recognizing the negative association between the two variables, a Plackett copula is used to formulate the joint distribution. The developed drought index, referred to as the drought management index (DMI), is able to differentiate the drought proneness of a given area when compared to other areas. An assessment of the sensitivity of the DMI to the length of the data segments used in evaluation indicates relative stability is achieved if the data segments are 5 years or longer. The proposed approach is illustrated with reference to the Malaprabha River basin in India, using four adjoining Climate Prediction Center grid cells of soil moisture data that cover an area of approximately $\mathrm{12,000}{\mathrm{km}}^2$.

Assessment and understanding of past climate is an important step for drought mitigation and water resources planning. In this study, streamflow simulated using the variable infiltration capacity (VIC) model was used for drought characterization for a time span of 1950–2000, and subsequently, regionalization was done for the state of Texas based on the annual drought severity level and drought duration. Droughts are regional in nature, and hence, identification of homogenous drought regions is important for investigating the drought characteristics within each of these regions. The concept of entropy was used for identification of homogenous regions based on drought severity and duration. A standardized version of mutual information, known as directional information transfer, was used for station grouping. The homogeneity of regions obtained was checked using L-moments. A total of eight regions were formed based on drought severity, and nine based on drought duration. Regions in west Texas were found to be critical in terms of severity, whereas east Texas showed the least severity. The longest drought duration was experienced in south Texas and lower valley zones, whereas the least drought duration was experienced in east Texas and the upper coast. Severely dry and extremely dry droughts were found to be restricted to the western and central parts of Texas.

The limited length of instrumental streamflow data impacts the true magnitude of natural interdecadal variability of water delivered from the Upper Green River Basin (UGRB). This limited period of instrumental record can be expanded by utilizing proxy records (reconstructed streamflow) derived from tree rings. Recent research has resulted in the development of nine streamflow reconstructions spatially located throughout the UGRB. This paper utilizes four of those nine reconstructed streamflow records and instrumental records to compare and analyze differences between the two streamflow records—human and natural. Three approaches were used for comparison and analysis: (1) Weibull distribution, (2) compound renewal, and (3) drought risk using bivariate probability distribution functions. This analysis has resulted in magnitude-duration-frequency curves for UGRB drought. Such probability curves and stochastic analysis can then be utilized in light of compact agreements and system storage to answer questions such as “How bad is it right now and what can we expect to happen next year?” This case study is intended to show statistical and observed differences between human (short-term) and natural (long-term) streamflow records and specifically target differences in long-term drought characteristics for this drainage basin.

In this paper, an integrated drought management scheme (IDMS) is developed for the Aharchay River Basin, located in northwestern Iran. This region is located in a semiarid region and has experienced considerable drought damage in recent years. Therefore, application of the proposed IDMS could greatly help to mitigate drought damage. The developed IDMS includes different tools and models needed for drought management. The utilized tools are classified into four groups, as follows: (1) data collection and validation, (2) rainfall-runoff prediction, (3) reservoir operation, and (4) drought analysis and monitoring. Future rainfall and runoff are projected using two computer models. Optimal reservoir operation is determined using a Bayesian stochastic genetic algorithm and Bayesian stochastic dynamic programming. These models are capable of incorporating new data of reservoir inflow and expert opinion in decision-making. For reservoir operation in drought conditions, hedging rules are also developed based on a mixed-integer nonlinear programming model. The hedging rules conserve water and adjust the operation in dry periods. Different drought indexes, including the standardized precipitation, surface water supply, and Palmer drought severity indexes, in addition to the hybrid drought index, which considers the socioeconomic aspects of drought, are used for drought evaluation. The socioeconomic aspects of drought are incorporated in drought management, utilizing an optimization model for land-use planning. The water pricing as a demand management tool for drought conditions is employed to measure drought impacts and usage efficiency of the limited available water resources. Based on drought characteristics in the study region, a drought warning scheme is also developed based on the drought monitoring data. To investigate the applicability of the proposed scheme, the future conditions of the basin hydroclimatic conditions are evaluated considering climate change impacts. To integrate these tools, a decision support system (DSS) package is developed, which includes different analytical modules needed for drought monitoring and management. The results show the significant value of integrated drought management schemes to identify and investigate the impacts of drought, and reduce the socioeconomic and environmental predicaments in the study region.

Global warming and its resulting climate change will affect different elements, such as water resources, in the future. One effect is that rainfall becomes very difficult to predict, as it is under the influence of several different elements. In this study, which considers Shahrekord synoptic station in Iran, various sources of uncertainty in rainfall prediction in the future and its effect on dry and wet spells are investigated. In the present research, CCSIRO, CGCM, ECHO-G, HADCM3, ECHAM, and PCM Atmospheric and Ocean General Circulation Model (AOGCM) models and ${\mathrm{A}}_1$, ${\mathrm{A}}_2$, ${\mathrm{B}}_1$, and ${\mathrm{B}}_2$ emission scenarios under three downscaling methods are examined. The results indicate a significant impact of the various downscaling methods on increasing the uncertainty band in rainfall estimation for the future. The AOGCM models in all of the scenarios except ${\mathrm{A}}_2$ are in agreement. The results of wet and dry spells estimation display a long-duration drought at the beginning of the upcoming 30-year period, followed by a long-duration wet spell.