Drought multistage prediction method based on state space and joint distribution

Abstract

The invention provides a drought multistage prediction method based on state space and joint distribution, and the method comprises the following steps: firstly extracting a drought index SPI time sequence, interpreting a state transition matrix of the drought index SPI time sequence through a first-order state space model, and calculating a drought state probability matrix of a to-be-predicted time period; then extracting a correlation factor of a previous time period of the time period to be predicted, constructing joint probability distribution of the correlation factor in each drought state and a drought index SPI time sequence, and deducing a conditional probability model of the drought index under the condition that the correlation factor is given; and finally, calculating the full-state distribution of the drought index of the to-be-predicted region in the to-be-predicted time period T, and based on the full-state distribution of the drought indexes, calculating to obtain a meanvalue and a 90% probability interval of the drought indexes. The invention has the beneficial effects that the state space method can reflect the drought uncertainty nature; influence factors are introduced through multiple prediction and classification, so that the drought prediction precision can be improved; and the process is considered more comprehensively and finely.

Description

technical field [0001] The invention relates to the technical field of drought prediction, in particular to a multi-level drought prediction method based on state space and joint distribution. Background technique [0002] In recent years, my country's Yunnan, Sichuan, Chongqing, Hunan, Hubei, Anhui and other provinces and cities have suffered severe droughts that occur once in 50 years or once in 100 years, causing incalculable economic losses. Researching a reliable drought prediction model and predicting the occurrence time and impact degree of drought events in advance will be of great significance to the sustainable development of society and economy. Currently, the most commonly used drought index is the standardized rainfall index (SPI), and the methods for forecasting the SPI index include time series methods, multivariate statistical regression methods, and machine learning methods. These models and theories are simple and feasible, but they have insufficient under...

AI Technical Summary

Problems solved by technology

[0003] In order to solve the technical problems of low prediction accuracy of drought index and less available information in the prior art, the present invention provides a multi-level drought prediction method based on state space and joint distribution, which mainly includes the following steps:

Images