Climate trends and variability at adaptation scale: Patterns and perceptions in an agricultural region of the Ethiopian Highlands

Analysis of climate variability and trends frequently takes place at large scale. For agricultural applications, however, highly localized climate conditions can be critically important. This certainly applies to tropical highland regions, where dissected topography and convectively dominated precipitation processes can lead to strong variability in both mean climate conditions and year-to-year climate variability. This study examines recent climate variability and trends (1981-2016) on Choke Mountain, located in the western Ethiopian Highlands. Through analysis of precipitation and temperature records at monitored locations, we explore observed variability in climate patterns and trends across sites and seasons. The lens for our spatial analysis is the agroecosystem (AES), defined on the basis of prevailing climate and cropping systems, which currently serve as the foundation for climate adaptation planning in the region. We find that interannual temperature variability is greatest in the hottest, driest AES, and is most pronounced in the dry season. All AES warmed significantly in all seasons over the analysis period, but the magnitude of trend was greatest in high elevation AES. Precipitation variability was also large across AES, with largest interannual variability found in the dry season. This season is frequently excluded in climate analyses, but it is a critical harvest time and irrigation period. Trends in rainfall anomaly and precipitation concentration index are less clear, but there is a tendency towards drying and increasing irregularity of rainfall. Interestingly, we find little association between the El Nino Southern Oscillation (ENSO) and temperature or precipitation variability at our study sites. This suggests that even though ENSO is a widely recognized driver of large-scale rainfall variability in the region, its impacts are highly spatially variable. This has implications for applying ENSO-based precipitation outlooks to agricultural management decisions. Farmer interviews reveal that local perceptions of climate variability and trends are generally consistent with the objective observations.