Spatiotemporal Trends of Climate Variability, Meteorological and Agricultural Droughts, and Their Effect on Land Surface Phenology in the Upper Gelana Watershed, Northeastern Highlands of Ethiopia
No Thumbnail Available
Date
2025-06-01
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Addis Ababa University
Abstract
Climate change remains a pressing global challenge, significantly impacting socio-ecological
systems and the livelihoods of millions, particularly in developing countries. In Ethiopia,
unpredictable climate patterns and extreme droughts have severely threatened rain-fed agriculture,
leaving many reliant on food assistance. As a result, understanding local climate variability, drought
characteristics, and their impacts is crucial for devising appropriate adaptation strategies to safeguard
smallholder farmers' livelihoods. This study explores climate variability and trends, meteorological
and agricultural droughts, spatial and temporal variations in land surface phenology (LSP), and
farmers' perceptions and adaptation strategies in response to climate variability and changeacross the
lower weina dega (LWD), upper weina dega (UWD), and dega agroecological zones (AEZs) of the
Upper Gelana watershed in the northeastern highlands of Ethiopia.
The first part of the study focusses on spatiotemporal climate variability and change across the
AEZs of the data-scarce Upper Gelana watershed, where rain-fed agriculture is the primary
livelihood source. We evaluated the performance of two widely used high-resolution satellite
precipitation datasets, Tropical Applications of Meteorology using SATellite and ground-based
observations (TAMSAT) and Climate Hazards Group Infrared Precipitation with Stations
(CHIRPS), using various categorical and continuous validation statistics in R. The spatiotemporal
variability was analyzed using TAMSAT rainfall and gridded temperature data from the Ethiopian
Meteorological Institute (EMI) with coefficient of variation (CV), precipitation concentration index
(PCI), and standardized rainfall anomaly (SRA). Trends were computed using the Mann-Kendall
(MK) test, Sen's slope estimator, and innovative trend analysis (ITA). The findings reveal
considerable inter-annual rainfall variability, with a significant positive trend in kiremt (main rainy
season) rainfall and slight declines in belg (short rainy season) rainfall. Decadal increases in kiremt
rainfall ranged from 96.1 to 104.8 mm, while belg rainfall declined by 14.0 to 16.4 mm across AEZs.
The minimum and maximum annual temperatures showed significant decreasing and increasing
trends, respectively, in LWD and UWD, while changes in dega were insignificant.
The second part investigates meteorological and agricultural droughts across the AEZs.
Meteorological drought analysis was conducted using the Standardized Precipitation Index (SPI)
with TAMSAT data, while agricultural drought was examined using Normalized Difference
Vegetation Index (NDVI), Land Surface Temperature (LST), Evapotranspiration (ET), and Potential
Evapotranspiration (PET). Drought characteristics such as intensity, severity, and duration were
analyzed using run theory in R, and spatial correlation analysis assessed the relationship between
meteorological and agricultural droughts. The findings indicate seasonal meteorological drought
variability, with kiremt droughts occurring more frequently than belg season droughts between 1991
and 2021. Meteorological droughts were detected in kiremt during 1991, 1993, 2002, 2009, and
ii
2015. Agricultural droughts, identified using ETDI and VHI, were more frequent during belg in
LWD and UWD (2002, 2008, 2009, 2011, 2012, 2013, and 2021) and in dega (2008, 2012, 2013, and
2015). Agricultural droughts were also observed during kiremt in 2002, 2008, and 2009 across all
AEZs. Pixel-wise correlation analysis revealed a statistically significant positive relationship
between meteorological and agricultural drought indices during both the belg and kiremt seasons,
indicating the potential translation of climatic stress to agricultural systems.
To assess the impact of climate change on vegetation ecosystems, we examined spatiotemporal
changes in LSP across AEZs and their association with climate variability and drought events from
2001 to 2021. MODIS NDVI (250-meter resolution) was used to compute phenological metrics,
including start of the season (SOS), end of the season (EOS), and length of the growing season
(LOS), as well as key ecosystem condition indicators such as base value, peak value, and amplitude
using MATLAB based TIMESAT software automated using Command Prompt (CMD) and R. Land
use was classified for 2001 and 2021 using the Random Forest algorithm, and areas with changes
were excluded from LSP trend detection to avoid misinterpretation. The relationship between LSP
changes and climate variables was assessed using pixel-wise pearson correlation and partial
correlation analyses. The results indicate that the dega AEZ experiences earlier SOS and longer LOS
compared to LWD and UWD. A delay in SOS and EOS was observed in 71.3% and 82% of the study
area, respectively, while LOS increased in nearly half of the area. The high spatial variability in the
peak and amplitude values suggests ecosystem condition variations across AEZs. SOS exhibited a
positive correlation with maximum temperature and a negative correlation with belg rainfall and
drought indices across large portions of the study area. Shorter LOS was associated with rising
temperatures, whereas increased rainfall extended LOS. The findings also indicate positive
correlations between drought indices and EOS in more than half of the study area. Partial correlation
analysis suggests that the negative impact of belg season meteorological drought on LSP is more
pronounced when translated into agricultural drought.
Lastly, we explored farmers perceptions of climate change, its impacts, and the determinants
influencing their adaptation choices using a multivariate probit model. The findings indicate that
farmers widely perceive rising temperatures, declining rainfall, and increased droughts, partially
aligning with statistical analyses. Adaptation strategies vary across AEZs: soil and water
conservation dominate in dega, while irrigation, crop variety changes, and planting adjustments are
prevalent in LWD and UWD. Education, extension services, and access to climate information
significantly influence adaptation choices.
Overall,the findings of the study revealed local scale spatiotemporal patterns of climate
variability, drought and their impact on LSP changes. Combined with the findings on perceived
impacts and existing adaptation practices, it will helpto raise awareness and contribute to the
development of AEZ-specific adaptation strategies to reduce the impacts of climate change on rain
fed agricultural dependent communities.
iii
Keywords: climate variability, agricultural drought, meteorological drought, land surface phenology,
remote sensing, perception, adaptation strategies
Description
Keywords
climate variability, agricultural drought, meteorological drought, land surface phenology, remote sensing, perception, adaptation strategies