Tadesse T., Demisse G.B., Zaitchik B., Dinku T.
National Drought Mitigation Center, University of Nebraska-Lincoln Nebraska, Lincoln, United States; Earth Science Department, Addis Ababa University, Addis Ababa, Piazza, Ethiopia; Department of Earth and Planetary Sciences, John Hopkins University, Baltimore, MD, United States; International Research Institute for Climate and Society, Columbia University, Palisades, NY, United States
Tadesse, T., National Drought Mitigation Center, University of Nebraska-Lincoln Nebraska, Lincoln, United States; Demisse, G.B., Earth Science Department, Addis Ababa University, Addis Ababa, Piazza, Ethiopia; Zaitchik, B., Department of Earth and Planetary Sciences, John Hopkins University, Baltimore, MD, United States; Dinku, T., International Research Institute for Climate and Society, Columbia University, Palisades, NY, United States
An experimental drought monitoring tool has been developed that predicts the vegetation condition (Vegetation Outlook) using a regression-tree technique at a monthly time step during the growing season in Eastern Africa. This prediction tool (VegOut-Ethiopia) is demonstrated for Ethiopia as a case study. VegOut-Ethiopia predicts the standardized values of the Normalized Difference Vegetation Index (NDVI) at multiple time steps (weeks to months into the future) based on analysis of "historical patterns" of satellite, climate, and oceanic data over historical records. The model underlying VegOut-Ethiopia capitalizes on historical climate-vegetation interactions and ocean-climate teleconnections (such as El Niño and the Southern Oscillation (ENSO)) expressed over the 24 year data record and also considers several environmental characteristics (e.g., land cover and elevation) that influence vegetation's response to weather conditions to produce 8 km maps that depict future general vegetation conditions. VegOut-Ethiopia could provide vegetation monitoring capabilities at local, national, and regional levels that can complement more traditional remote sensing-based approaches that monitor "current" vegetation conditions. The preliminary results of this case study showed that the models were able to predict the vegetation stress (both spatial extent and severity) in drought years 1-3 months ahead during the growing season in Ethiopia. The correlation coefficients between the predicted and satellite-observed vegetation condition range from 0.50 to 0.90. Based on the lessons learned from past research activities and emerging experimental forecast models, future studies are recommended that could help Eastern Africa in advancing knowledge of climate, remote sensing, hydrology, and water resources. Key Points Developed new satellite-based prediction model called VegOut-Ethiopia Demonstrated application of VegOut-Ethiopia model to a recent drought year Highlighted future research opportunities under evolving climate conditions © 2014. American Geophysical Union. All Rights Reserved.
Atmospheric pressure; Climatology; Drought; Forecasting; Remote sensing; Research; Satellites; Vegetation; Water resources; Correlation coefficient; Drought monitoring; Environmental characteristic; Normalized difference vegetation index; Research opportunities; Southern oscillation; Vegetation condition; Vegetation monitoring; Climate models