Table 1 Overview of key drought indices across meteorological, agricultural, hydrological, socioeconomic, and environmental dimensions; summarizes each drought dimension’s principal indicators, highlighting their fundamental analytical aspects when applied as single indices and the inherent limitations if used in isolation.
Drought dimension | Key indicators | Perspective on single-index drought analysis aspects | Disadvantages of single-index drought analysis if used alone |
|---|---|---|---|
Meteorological | Palmer Drought Severity Index (PDSI)14 | Integrates precipitation and temperature over time using a water-balance model to detect prolonged drought conditions and provide historical drought severity classification at monthly to annual scales. Sensitive to prolonged soil moisture deficiency and useful for tracking multi-seasonal meteorological droughts | Limited applicability in regions lacking reliable soil moisture or where evapotranspiration estimation is uncertain. Does not account for rapid drought onset (“flash droughts”), and calibration is region-dependent, which can hinder spatial comparability. May lag real-time drought evolution due to its smoothing characteristics |
Standardized Precipitation Index (SPI)15 | Quantifies precipitation deficit relative to the long-term mean at various accumulation periods (e.g., 1-, 3-, 12-month), enabling the detection of short- and long-term meteorological droughts with high temporal flexibility and statistical comparability between regions and periods | Ignores temperature and evapotranspiration, thus underestimating drought severity in warming climates or during heat-driven droughts. Fails to capture the role of changing potential evapotranspiration, possibly misrepresenting actual water deficits, especially under climate change | |
Standardized Precipitation Evapotranspiration Index (SPEI)16 | Extends SPI by incorporating both precipitation and potential evapotranspiration, allowing assessment of how combined deficits and increased atmospheric demand shape meteorological drought severity. Highly responsive to warming-induced drought intensification | Sensitive to uncertainties in evapotranspiration estimation (method-dependent), and may be less stable in arid environments where evapotranspiration approaches precipitation. Assumes standardized input data; affected by errors in temperature or PET records | |
Standardized Precipitation Anomaly Index (SPAI)17 | Calculates precipitation anomalies relative to the long-term climatological mean and standardizes these for probabilistic drought severity classification; robust for strongly seasonal or periodic climates and effective for inter-regional comparisons | Does not account for atmospheric demand (temperature, PET), possibly underestimating drought risk in the presence of concurrent warming. Can be less sensitive to multi-factor drought drivers and may miss drought propagation into soil or hydrology if used alone | |
Agricultural | Crop Moisture Index (CMI)18 | Quantifies short-term soil moisture conditions in the root zone, using precipitation and evapotranspiration data, to detect rapidly developing agricultural drought. Sensitive to recent weather and calibrated for crop response during the growing season | Limited to short-term fluctuations; does not capture persistent, long-term drought or deep soil moisture deficits. Regionally calibrated and may not perform well in diverse climates. Insensitive to non-climatic crop stressors (e.g., pests, management practices) and ignores vegetative response |
Crop Water Stress Index (CWSI)19 | Remotely sensed or field-based index measuring canopy temperature relative to air temperature to assess plant water stress and transpiration deficit at the field to landscape scale, especially effective under clear sky conditions | Highly sensitive to atmospheric and radiative conditions; confounded by factors other than water stress (e.g., disease, phenology). Provides instantaneous assessments but does not reflect cumulative or soil-root zone drought. Not suitable for long-term or spatially heterogeneous regions without extensive validation | |
Relative Water Deficit (RWD)20 | Expresses the relative reduction in plant water content compared to maximum capacity, directly reflecting the physiological moisture status of crops and their immediate drought response. Valuable for linking field observations to agronomic drought effects | Measurements are labor-intensive and spatially discontinuous. Susceptible to local soil and crop variability; not amenable to regional or long-term monitoring without scaling. Ignores meteorological and hydrological context, missing external drought propagation | |
Vegetation Condition Index (VCI)21 | Derived from NDVI time series, reflecting vegetation greenness relative to historical extremes, allowing spatially explicit detection of crop stress and early agricultural drought warning using satellite data | Sensitive to non-drought drivers of NDVI change (e.g., land use, crop rotation, pest outbreaks). May lag early physiological drought if vegetation is slow to respond. Does not differentiate between moisture and thermal stress unless integrated with other indices | |
Vegetation Health Index (VHI)22 | Integrates VCI (greenness) and TCI (thermal stress) for comprehensive assessment of plant health and combined response to moisture and temperature anomalies, effective for large-scale, near-real-time agricultural drought monitoring | Affected by cloud contamination, atmospheric noise, and remote sensing data gaps. Still limited in attributing crop stress solely to drought, as disease, fertility, or management can also impact VHI. Provides a generalized rather than crop-specific drought diagnosis, and may not capture sub-seasonal dynamics in rapidly changing environments | |
Hydrological | Palmer Hydrological Drought Index (PHDI)23 | Integrates long-term streamflow and soil moisture conditions to identify hydrological drought intensity and duration, providing standardized monthly drought severity classifications. Useful for extended surface water supply assessment and resource planning | Sensitive to calibration parameters and soil moisture data quality; regional applicability varies due to soil and hydrology differences. It may not capture short-term hydrological variability or groundwater droughts effectively. Potential lag in detecting drought onset and recovery |
Surface Water Supply Index (SWSI)24 | Combines streamflow, reservoir storage, and snowpack data to quantify overall surface water availability. Provides a comprehensive monthly drought severity metric sensitive to multiple water sources used in reservoir management | Data-intensive, requiring accurate reservoir and snowpack measurements that may not be readily available in all regions. Limited applicability outside snow-influenced basins and less effective in purely rainfed systems. May mask sector-specific vulnerabilities due to its composite nature | |
Standardized Runoff/Streamflow Index (SRI/SSI)25 | Standardizes streamflow data using statistical fitting (e.g., gamma distribution) to characterize drought severity and duration. Effective for basin-scale hydrological drought monitoring with monthly time resolution | Dependent on the quality and continuity of hydrological data, especially in data-scarce regions. Does not account for groundwater or ecohydrological droughts independently. It may not fully represent complex hydrological drought dynamics in regulated or highly managed basins | |
Socioeconomic | Multivariate Standardized Reliability and Resilience Index (MSRRI)26 | Quantifies socioeconomic drought by integrating climate-driven supply variability and water system reliability and resilience, using water availability, demand, and system performance metrics. Useful for linking drought to household and regional water security | May miss sector-specific or subsystem vulnerabilities (e.g., agriculture vs. domestic supply). Relies heavily on system operation/management data, which may not be available or accurate. Does not explicitly capture non-supply-related impacts (social resilience, economic loss) if used alone |
Improved MSRRI (IMMSRI)27 | Advances MSRRI by adding socioeconomic and resilience adjustment factors, and includes multidimensional water system characteristics (e.g., infrastructure, governance). Offers improved assessment of adaptability during drought | Results are highly sensitive to weighting and normalization approaches, which may vary across regions. Can mask abrupt changes at the subsystem or community scales due to aggregate scoring. Ignores feedback from environmental and agricultural drought unless used in a composite framework | |
Socioeconomic Drought Index (SEDI)28 | Integrates both water availability and socioeconomic responses (e.g., population, GDP, consumption) to reflect human and economic vulnerability during drought events. Captures direct human impacts of hydrological stress | Prone to data quality and availability constraints, especially for high-frequency socioeconomic data. May oversimplify the interaction between climate variability and human adaptive behavior, and does not account for natural ecosystem impacts, thus providing a partial view if used in isolation | |
Water Resources System Resilience Index (WRSRI)29 | Measures the resilience of water supply systems under stress by quantifying the frequency, severity, and recoverability of system failures. Focused on infrastructure/system preparedness and recovery | Relies on the modeling of system recovery, which can be uncertain or poorly calibrated. Provides limited understanding of underlying drought causality or societal impacts beyond system performance. Neglects hydrological, agricultural, and ecological drought propagation if used as the sole indicator | |
Standardized Water Supply and Demand Index (SWSDI)30 | Evaluates socioeconomic drought by standardizing the deficit or surplus between water supply and demand at various spatial and temporal scales; useful for real-time drought risk communication and water management | Requires robust, timely data on both supply and demand-often a challenge in data-limited settings. Does not differentiate between causes of supply deficits (e.g., climate vs. management failure). Ignores cascading drought impacts outside direct supply–demand mismatch, such as ecological or food system vulnerabilities | |
Environmental | Environmental Drought Index (EDI)31 | Quantifies environmental drought by capturing periods when river flows fall below ecologically defined thresholds, integrating both drought duration and water deficit severity (water shortage and persistence) to signal ecosystem stress. Provides an explicit, sector-specific assessment of ecological drought, accounting for minimum flows critical to sustaining aquatic and riparian habitats. Enables early warning for unmanaged systems and highlights chronic or acute ecological vulnerability that may not be reflected in meteorological, agricultural, or hydrological indices | Restricted to ecohydrological flow thresholds and may not reflect broader socioeconomic, agricultural, or hydrological drought propagation events. Sensitive to the accuracy of minimum flow estimations and threshold selection, and may underestimate ecosystem risk if ecological water needs are poorly defined or regionally unvalidated. Does not capture cross-sectoral interactions or the cascading effects of drought across other systems unless used in conjunction with complementary indices |
