Standardized Precipitation Index (SPI)

The Standardized Precipitation Index (SPI) is a probability index that gives a better representation of abnormal wetness and dryness than the Palmer Severe Drought Index (PSDI). The World Meteorological Organization (WMO) recommends, that all national meteorological and hydrological services should use the SPI for monitoring of dry spells. Some advantages of the SPI

• It requires only monthly precipitation.
• It can be compared across regions with markedly different climates.
• The standardization of the SPI allows the index to determine the rarity of a current drought.
• It can be created for differing periods of 1-to-36 months.

The SPI is obtained by fitting a gamma or a Pearson Type III distribution to monthly precipitation values. The current implementation of dim_spi_n uses a 2-parameter gamma distribution fit where the shape and scale parameters are maximum liklihood estimates as described in

        A Note on the Gamma Distribution
        Thom (1958): Monthly Weather Review, pp 117-122.
                     specifically: eqn 22 for gamma; just above eqn 21

However, there is some variation in the methods used to derive the SPI. Generally, monthly precipitation is not normally distributed so a transformation is performed such that the derived SPI values follow a normal distribution. The SPI is the number of standard deviations that the observed value would deviate from the long-term mean, for a normally distributed random variable. One interpretation of the resultant values is:

         [+,-]2.00 and above/below: exceptionally [wet,dry] 
         [+,-]1.60 to 1.99: extremely [wet,dry]
         [+,-]1.30 to 1.59: severely [wet,dry] 
         [+,-]0.80 to 1.29: moderately [wet,dry] 
         [+,-]0.51 to 0.79: abnormally [wet,dry] 
         [+,-]0.50:  near normal

Explanation of the SPI at different lengths and sample spatial pattterns over the USA at different run times are available.

More information can be obtained at the ClimateDataGuide.

As noted in the dim_spi_n documentation, at least 30 years of data are recommended.