Drought Forecasting Using the Standardized Precipitation Index

University of Catania, Catania, Sicily, Italy
Water Resources Management (Impact Factor: 2.6). 05/2007; 21(5):801-819. DOI: 10.1007/s11269-006-9062-y


Unlike other natural disasters, drought events evolve slowly in time and their impacts generally span a long period of time.
Such features do make possible a more effective drought mitigation of the most adverse effects, provided a timely monitoring
of an incoming drought is available.

Among the several proposed drought monitoring indices, the Standardized Precipitation Index (SPI) has found widespread application
for describing and comparing droughts among different time periods and regions with different climatic conditions. However,
limited efforts have been made to analyze the role of the SPI for drought forecasting.

The aim of the paper is to provide two methodologies for the seasonal forecasting of SPI, under the hypothesis of uncorrelated
and normally distributed monthly precipitation aggregated at various time scales k. In the first methodology, the auto-covariance matrix of SPI values is analytically derived, as a function of the statistics
of the underlying monthly precipitation process, in order to compute the transition probabilities from a current drought condition
to another in the future. The proposed analytical approach appears particularly valuable from a practical stand point in light
of the difficulties of applying a frequency approach due to the limited number of transitions generally observed even on relatively
long SPI records. Also, an analysis of the applicability of a Markov chain model has revealed the inadequacy of such an approach,
since it leads to significant errors in the transition probability as shown in the paper. In the second methodology, SPI forecasts
at a generic time horizon M are analytically determined, in terms of conditional expectation, as a function of past values of monthly precipitation.
Forecasting accuracy is estimated through an expression of the Mean Square Error, which allows one to derive confidence intervals
of prediction. Validation of the derived expressions is carried out by comparing theoretical forecasts and observed SPI values
by means of a moving window technique. Results seem to confirm the reliability of the proposed methodologies, which therefore
can find useful application within a drought monitoring system.

Download full-text


Available from: Brunella Bonaccorso,
    • "These two indexes were used in this study to analyze the pattern of precipitation in the southwest China. Precipitation extremes are in close relation with flood events and droughts (Cancelliere et al. 2007; Jiang et al. 2013b; Kundzewicz et al. 2010; Tsakiris and Vangelis 2004; Zhao et al. 2012), therefore, a comparative analysis of these extreme precipitation indices is of scientific and practical merits for better understanding of flood events in the study area. Another concern for precipitation extreme change is the detection of possible presence of long memory in the data, e.g., if the change has occurred, whether and how the change persists. "
    [Show abstract] [Hide abstract]
    ABSTRACT: Changes in the spatiotemporal patterns of precipitation have great impacts on drought/flood risk and utilization of water resources. In this study, we presented the results of a comparative analysis of spatial-temporal variability of precipitation in the Southwest China. The analysis investigated the trends of annual precipitation and explored the changes of two indices: the precipitation concentration index (PCI) and the concentration index (CI) which are designed for measuring seasonality and daily heterogeneity, respectively. The trends of annual precipitation and CI were tested by the Mann-Kendall method. The results show a significant seasonality of the rainfall distribution and very inhomogeneous temporal distribution of the daily rainfall in the study area. Positive trends in the CI were found at most stations, although most of them were not statistically significant. To detect the futures trends of precipitation in the study area, Hurst’s rescaled range (R/S) analysis was introduced and the corresponding Hurst Exponent was estimated. The results suggested that some drought hazards will happen in the intersection of Sichuan, Guizhou and Yunnan, and the west part of Sichuan, the north part of Chongqing and the middle part of Yunnan are under the risk of flood in the future.
    Water Resources Management 09/2015; 29(11). DOI:10.1007/s11269-015-1038-3 · 2.60 Impact Factor
  • Source
    • "Meanwhile, the TCCs of the winter corrected NAO are 0.46 and 0.54 based on ECMWF and CNRM, compared with TCCs of 0.34 and 0.46 based on ECMWF and CNRM for the Scheme-I, respectively (see Table 1 and Table 2). In order to accurately assess predictive skill of the Scheme-II for the two coupled models of DEMETER, the 95% prediction intervals are added (Cancelliere et al., 2007), and almost all the observed NAO index fall within the predictive range, suggesting better prediction skill (Figure 7). This also illustrates that the prediction skill of Scheme-II is higher than Scheme-I for both ECMWF and CNRM based two NAO indices, which could be attributed to the NAO/SST association included in Scheme-II. "
    [Show abstract] [Hide abstract]
    ABSTRACT: The winter North Atlantic Oscillation (NAO) is a crucial part of our understanding of Eurasian and Atlantic climate variability and predictability. In this study, we developed effective prediction schemes based on the interannual increment prediction method and verified their performance based on the climate hindcasts of the coupled ocean–atmosphere climate models. This approach utilizes the year-to-year increment of a variable (i.e. a difference in a variable between the current year and the previous year, e.g. DY of a variable) as the predictand rather than the anomaly of the variable. The results demonstrate that the new schemes can generally improve prediction skill of the winter NAO compared to the raw coupled model's output. Also, the new schemes show higher skill in prediction of abnormal NAO cases than the climatological prediction. Scheme-I uses just the NAO in the form of year-to-year increments as a predictor that is derived from the direct outputs of the models. Scheme-II is a hybrid prediction model that contains two predictors: the NAO derived from the coupled models and the observed preceding autumn Atlantic sea surface temperature in the form of year-to-year increments. Scheme-II shows an even better prediction skill of the winter NAO than Scheme-I.
    International Journal of Climatology 04/2015; DOI:10.1002/joc.4330 · 3.16 Impact Factor
  • Source
    • "In the current study, a comprehensive review was carried out using the SPI in 157 previous studies (from 1998 till early 2013), and it became clear that the 1-, 3-, 6-, 9-, 12-, and 24-months' time scales have been used 65, 97, 89, 46, 96, 42 times, respectively. However, most researchers and users may not have already had any idea of the time scale being used in their works, or they may have simply used a hypothetical short or long time scale to represent a certain statistical application in the drought studies; for instance, analyzing the transition probability of drought classes on the 12-month time scale (Paulo et al. 2005; Paulo and Pereira 2007; Paulo and Pereira 2008) or forecasting drought on the scales of 3, 6, 9, 12, and 24 months (Mishra and Desai 2005; Cancelliere et al. 2007). Note that unawareness of the appropriate time scale and calculating the drought index on several time scales may cause confusion when interpreting the results. "

Show more