Atmospheric drivers affect crop yields in Mozambique

Abstract

Climate change has been inducing variations in the statistics of both the large-scale weather patterns and the local weather in many regions of the world, and these variations have been affecting several human activities, including agriculture. In this study, we look at the links between large-scale weather patterns and local weather as well as agriculture, with a specific regional focus on Mozambique between 1981 and 2019. First, we investigated linear trends and links between large-scale weather patterns and local weather in the region using the ERA5 dataset. We used the same data to investigate how climate change has been affecting the statistics of large-scale weather patterns. Then, we derived Mozambique country-level cereal yield data from FAO and linked it up with climate and weather data to assess what is the relationship between large-scale patterns and local agronomic outputs using a multiple linear regression (MLR) model with crop yield as the response variable and climate drivers as predictors. The results indicate that in Mozambique, the crop season warmed substantially and consistently with climate change-induced global warming, and the rainy season had become drier and shorter, with precipitation concentrated in fewer, more intense events. These changes in the local weather have been linked to variations in the statistics of large-scale weather patterns that characterize the (large-scale) atmospheric flow over the region. Our results indicate a negative impact on yield associated with climate change, with average yield losses of 20% for rice and 8% for maize over the analyzed period (1981–2019). This negative impact suggests that, at the country scale, further future warming during the growing season may offset some of the cereal yield gains from technological advances.