Standardized Precipitation Index
- Introduced by
- Thomas McKee, Neil Doesken, and John Kleist
- Year
- 1993
- Type
- Probabilistic drought indicator
- Family
- Drought and precipitation analysis
What does it do?
The Standardized Precipitation Index (SPI) is a climate index that quantifies precipitation anomalies relative to historical norms, standardized to account for differences in precipitation climatology across regions.
Introduced by McKee, Doesken, and Kleist in 1993, SPI has become a primary tool for drought detection and characterization, adopted by meteorological agencies worldwide for operational drought monitoring and early warning systems.
The intuition behind the method
SPI measures how many standard deviations away from the long-term mean the cumulative precipitation over a specified time period (1, 3, 6, 12, or 24 months) lies. Negative SPI values indicate precipitation deficits (dry conditions); positive values indicate surpluses (wet conditions). By using standardized values, SPI allows direct comparison across different climate zones with very different mean precipitation patterns. A SPI of -1 represents a moderate drought; -2 represents a severe drought. The beauty of SPI is its simplicity: only precipitation data are needed, making it applicable globally even where other meteorological variables are unavailable.
History
Introduced at the Eighth Conference on Applied Climatology in 1993, SPI filled a critical gap in drought monitoring by providing a mathematically standardized, spatially comparable index. It was adopted by the World Meteorological Organization in 2012 as a recommended drought indicator. The WMO publication 'Standardized Precipitation Index User Guidebook' (2012) codified best practices. Recent developments include high-resolution gridded SPI from satellite precipitation, machine-learning approaches to refine distribution assumptions, and coupled SPI-SPEI frameworks for comprehensive drought assessment.
When to use it
Use SPI to quantify and monitor drought conditions, particularly for agricultural and water resource management applications. SPI is ideal for detecting onset and intensity of droughts at monthly to inter-annual timescales. Combine SPI with other indices (SPEI, NDVI) for comprehensive drought characterization. SPI is less useful in tropical regions with high-intensity convective rainfall or in areas with snow accumulation where precipitation is not immediately hydrologically effective.
Limitations and considerations
- Based solely on precipitation; does not account for temperature, evapotranspiration, or soil moisture, which influence actual water availability
- Probability distribution assumptions (gamma, Pearson III) may not hold at all locations or time periods, particularly in regions with high precipitation variability
- Requires long historical records (30+ years) to establish robust climatology; unreliable in data-short regions or for new weather stations
- Short accumulation periods (1–3 months) can be noisy and may not reflect true hydrological drought conditions
Source conditions do not replace independent scientific review or validation by an analysis engine.
Strengths
- Data-parsimonious: requires only precipitation data, enabling application in data-sparse regions
- Standardized across regions, allowing direct comparison of drought intensity between different climate zones
- Flexible timescale: SPI can be computed at 1-, 3-, 6-, or 12-month intervals to capture droughts of different durations
- Widely adopted and operationally used, ensuring compatibility with existing drought monitoring networks and decision-making frameworks
Applications
- Operational drought monitoring: national meteorological agencies use SPI to issue drought warnings, trigger water restrictions, and guide emergency response
- Agricultural decision-making: farmers and irrigation districts use SPI to optimize water allocation, schedule planting, and adjust crop selection
- Hydropower reservoir management: dam operators monitor SPI to forecast inflows and plan generation schedules across seasons and years
- Climate change impact assessment: SPI trends are used to project future drought frequency and intensity under different climate scenarios
How it works
How it works
Long-term monthly precipitation observations are collected from weather stations, satellite data, or reanalysis products. A minimum of 30 years of data is recommended to establish robust climatological statistics.
Precipitation data are quality-controlled to detect and remove erroneous values, duplicates, and gaps. Missing data are either filled via interpolation or months are excluded if gaps exceed acceptable thresholds.
For each accumulation period (1, 3, 6, 12, 24 months), rolling sums of precipitation are calculated. The probability distribution of these accumulations is fitted (typically using a gamma or Pearson Type III distribution).
The cumulative distribution function is computed for each accumulation period. Rolling precipitation sums are normalized using the fitted distribution to produce standardized anomalies (Z-scores).
SPI values are categorized into drought classes (near-normal, mild, moderate, severe, extreme) and compared against historical thresholds. Spatial and temporal patterns of SPI anomalies reveal drought extent and duration.
Common pitfalls
- Confusing SPI with actual water deficit; SPI is a probability-based index and does not directly quantify hydrological impacts
- Selecting inappropriate accumulation periods for the application (e.g., 1-month SPI for agricultural drought when 3- or 6-month SPI is more relevant)
- Over-relying on SPI without integrating soil moisture, streamflow, or groundwater data to understand actual water availability
Frequently asked questions
What is the difference between SPI and SPEI?
SPI uses only precipitation, while SPEI (Standardized Precipitation-Evapotranspiration Index) also incorporates temperature and evapotranspiration. SPEI better captures the combined effect of precipitation deficit and increased evaporative demand, making it more physically relevant for soil moisture and hydrological drought. However, SPEI requires additional temperature data.
How is a SPI value interpreted?
A SPI of 0 represents the long-term mean. Positive values (up to +3 or more) represent wet conditions; negative values (down to -3 or lower) represent dry conditions. A commonly used classification: SPI > 1.5 is extremely wet; 1.5 to 1.0 is very wet; 1.0 to 0.5 is moderately wet; -0.5 to 0.5 is near normal; -0.5 to -1.0 is moderately dry; -1.0 to -1.5 is severe drought; -1.5 to -2.0 is very severe drought; < -2.0 is extreme drought.
Why use multiple timescale SPI values?
Different timescales capture droughts affecting different sectors. 1-month SPI shows rapid changes and is sensitive to short-term rainfall events. 3- and 6-month SPI reflect soil moisture and agricultural drought. 12-month SPI indicates hydrological drought affecting streamflow and groundwater. Using all timescales together provides a complete picture of drought at all relevant timescales.
Can SPI be used for precipitation forecasting?
No. SPI is a retrospective index based on observed precipitation. To forecast future drought, one must first forecast precipitation (using weather or climate models) and then compute SPI from the forecasted precipitation. SPI itself is not predictive.
Related methods
Sources
McKee, T. B., Doesken, N. J., & Kleist, J. (1993). The relationship of drought frequency and duration to time scales. Proceedings of the Eighth Conference on Applied Climatology, 179-184.
Open publication referenceCopied.Lloyd-Hughes, B., & Saunders, M. A. (2002). A drought climatology for Europe. International Journal of Climatology, 22(13), 1571-1592.
Open DOI referenceCopied.