气候变化领域集成服务门户(CCPortal): Exploring improvements to the design of an operational seasonal forage scarcity index from NDVI time series for livestock insurance in East Africa

CCPortal

DOI	10.1016/j.jag.2019.05.018
	Exploring improvements to the design of an operational seasonal forage scarcity index from NDVI time series for livestock insurance in East Africa
	De Oto L.; Vrieling A.; Fava F.; de Bie K.C.A.J.M.
发表日期	2019
ISSN	15698432
卷号	82
英文摘要	Recurrent drought represents a major threat in arid and semi-arid regions of East Africa where pastoralists depend on their livestock for subsistence. In Kenya and southern Ethiopia, an existing satellite-based index insurance scheme aims to protect pastoralists against the adverse effects of drought. Under that scheme, payouts are made based on an area-aggregated seasonal forage scarcity index derived from remotely-sensed Normalized Difference Vegetation Index (NDVI). NDVI values are directly averaged per unit areas of insurance (UAI), which are based on administrative borders but take into limited account the ecological variability within the unit. The choice of administrative boundaries at the onset of the analysis may negatively impact the performance of the product. Our study explores an alternative index design based on an ecological stratification of the study area. First, we performed an unsupervised classification of NDVI time series from the Moderate Resolution Imaging Spectroradiometer (MODIS) to group pixels with similar temporal NDVI trajectories. Next, we used average NDVI profiles and ancillary data to discard areas deemed insignificant for forage production. We then transformed NDVI values into z-scores to assess how each pixel relates to the multi-year distribution of NDVI values per class and season. In the final step, we calculated the alternative forage scarcity index, which is the percentage of pixels with anomalously low NDVI values per season and UAI (i.e. z-score ≤ −1.0). To evaluate its performance, we compared unit-level results for both the original and alternative designs against spatially-aggregated monthly household survey data on livestock mortality from 16 sample sites corresponding to eight administrative units within the study area. Besides performing better in predicting livestock mortality (i.e. increases of 53% and 39% in correlation coefficients as measured by Pearson's r and Spearman's ρ, respectively) and strengthening the ecological significance of the index, the proposed design has a number of other advantages: 1) the index is calculated using a much larger statistical basis, 2) it allows for analysis of forage conditions at sub-unit level, and 3) it offers a more flexible structure for payout calculation. These advantages could be of particular relevance for expanding the index-insurance scheme to agro-pastoral regions characterized by more heterogeneous landscapes. Finally, we propose to consider using this approach to better account for the ecological variability of rangelands in other NDVI-based early warning and monitoring systems. © 2019 Elsevier B.V.
英文关键词	Drought; Index insurance; ISODATA clustering; MODIS; NDVI time series; Phenology
语种	英语
scopus关键词	algorithm; design; drought; forage; insurance system; livestock farming; MODIS; NDVI; phenology; seasonal variation; time series analysis; unsupervised classification; Ethiopia; Kenya
来源期刊	International Journal of Applied Earth Observation and Geoinformation
文献类型	期刊论文
条目标识符	http://gcip.llas.ac.cn/handle/2XKMVOVA/156445
作者单位	University of Twente, Faculty of Geo-information Science and Earth Observation, P.O. Box 217, Enschede, AE 7500, Netherlands; International Livestock Research Institute, P.O. Box 30709, Nairobi, 00100, Kenya
推荐引用方式 GB/T 7714	De Oto L.,Vrieling A.,Fava F.,et al. Exploring improvements to the design of an operational seasonal forage scarcity index from NDVI time series for livestock insurance in East Africa[J],2019,82.
APA	De Oto L.,Vrieling A.,Fava F.,&de Bie K.C.A.J.M..(2019).Exploring improvements to the design of an operational seasonal forage scarcity index from NDVI time series for livestock insurance in East Africa.International Journal of Applied Earth Observation and Geoinformation,82.
MLA	De Oto L.,et al."Exploring improvements to the design of an operational seasonal forage scarcity index from NDVI time series for livestock insurance in East Africa".International Journal of Applied Earth Observation and Geoinformation 82(2019).