Linkages between common wheat yields and climate in Morocco (1982-2008)

In Morocco, wheat production shows a high inter-annual variability due to uncertain rainfall. In view of the importance of this resource to the country's economy, it is important to gain a better understanding of the natural large-scale climate oscillation governing this variability. In this study, we analyzed de-trended (1) time series of common wheat yields (1983-2008) from 11 agricultural provinces that account for 80 % of national wheat production; (2) monthly rainfall and 10-day temperature from ten meteorological stations; (3) 10-day normalized difference vegetation index (NDVI) from the AVHRR sensor; (4) monthly atmospheric climate indices [North Atlantic Oscillation (NAO) and Scandinavian Pattern (SCA)] and monthly 500 hPa geopotentials fields; and (5) monthly sea surface temperature (SST) fields and indices (NIAO3, Tropical North Atlantic and Tropical South Atlantic). The relationship between rainfall and temperature during tillering in early winter and grain filling in early spring and wheat yields already observed at the plot scale was also found to be significant at the provincial scale. The linkages between wheat yields and large scale climate have been analyzed for the first time over Morocco. In agreement with previous studies, results show a complex and competing influence of different climate phenomena. The NAO is found to be significantly related to yields during the early stage of wheat growth in December, whereas the SCA correlates with yields later in the season, in January and February. Interesting lagged correlations with higher lead time are also highlighted, with the leading modes of SST variability in the equatorial Atlantic during October (the "Atlantic Nio" mode) and in the North Atlantic (the "Atlantic tripole" mode) in February. Our conclusion is that regional climate indices and variables represent valuable information with which to increase lead time and skill regarding wheat yield predictions in Morocco.