Standardised Precipitation Index (SPI)

A number of drought/precipitation indices are currently under investigation at CIMH, with the view to them being used together to monitor drought and periods with extremely high precipitation. The Standardised Precipitation Index (SPI) is one such index under investigation

The Standardised Precipitation Index (SPI), developed by T.B. Mckee, N.J. Doesken and J. Kleist (McKee et al. 1993) of Colorado State University is an index that, if used carefully, can provide early warning of an extended drought period and aid in assessing drought severity. It can also provide similar information at the other end of the spectrum- extremely high precipitation. SPI is basically a representation of rainfall in units of standard deviation. Positive values indicate greater than median rainfall; negative values indicate less than median rainfall. The severity classes are shown in Table 1.

SPI Values and precipitation intensities (Mckee et al 1993)
SPICategoryProbability (%)
2.0 +Extremely wet2.3
1.5 to 1.99Very wet4.4
1.0 to 1.49Moderately wet9.2
-0.99 to 0.99Near normal68.2
-1.0 to -1.49Moderately dry9.2
-1.5 to -1.99Severely dry4.4
-2.0 and lessExtremely dry2.3

Impacts experienced (particularly in agriculture) during the Caribbean drought of 2009-2010, prompted a re-thinking of the categories from the traditional table above; similar to that currently used by US Drought Monitor. The categories in Table 2. below are being used since 2011.

Table SPI classification used from January 2011
SPI ValueCategorySPI ValueCategory
-0.50 to -0.01Normal0.50 to 0.01Normal
-0.80 to -0.51Slightly dry0.80 to 0.51Slightly wet
-1.30 to -0.81Moderately dry1.30 to 0.81Moderately wet
-1.60 to -1.31Severely dry1.60 to 1.31Very wet
-2.00 to -1.61Extremely dry2.00 to 1.61Extremely wet
less than or equal to -2.01Exceptionally drygreater than or equal to +2.01Exceptionally wet

The SPI is flexible and can be calculated for different time scales. A time scale analysis reflects the impact of drought on the availability of the different water resources. A one-month SPI analysis reflects short term soil moisture and crop stress especially during the growing season. A three-month SPI analysis reflects short to medium term moisture and can give an indication of available moisture conditions at the beginning of the growing season. A six-month SPI analysis reflects medium term trends in rainfall and is effective in showing rainfall distribution over distinct seasons as well as being associated with anomalous stream flows and reservoir levels, which takes longer to manifest itself than does agricultural drought. A twelve-month SPI can indicate potential periods of shortfall in groundwater amounts.

Using NCEP/NCAR reanalysis data in combination with data from land based stations, SPIs are calculated in 1, 3, 6 and 12 month time intervals for the Caribbean basin as part of the Caribbean Drought and Precipitation monitoring Network. CIMH hopes to add more land stations in due course.

When used in combination with forecast products, such as seasonal precipitation forecasts, projections can be made on the future status of drought (or extreme high rainfall) for the region. CIMH is currently investigating the use of other indices to monitor drought and precipitation.