Agroclimatic Modeling of the Drought Trend in the State of Sinaloa, Mexico, Using the PDO--AMO Indices.

The goal was to model the trend of meteorological droughts (MeD) using the Pacific decadal oscillation (PDO) and Atlantic multidecadal oscillation (AMO) indices. PDO-AMO series were obtained from the National Oceanic and Atmospheric Administration. In 12 weather stations in the state of Sinaloa (1981-2017), the agricultural standardized precipitation index (aSPI) and reconnaissance drought index (RDI) were calculated. The linear (SLT) and non-parametric (SNT) significant trends of the aSPI and RDI were calculated. A principal component analysis was applied to SLT-SNT and the first observed principal component (Z PC-1o) was extracted. The first calculated principal component was modeled through a linear regression of Z PC-1c (dependent variable) on PDO-AMO (independent variables). The correlations between Z PC-1o vs Z PC-1c = 0.522 and the linear trend of Z PC-1c = 0.501, were significantly different from zero. This study contributes to addressing a research gap not otherwise explored to date in Sinaloa: modeling of the trend in MeD through aSPI-RDI and PDO-AMO. The model can be used to help schedule agricultural irrigation at the most productive times. [ABSTRACT FROM AUTHOR]