Abstract
Image time series of high temporal and spatial resolution capture land surface dynamics of heterogeneous landscapes. We applied the ESTARFM (Enhanced Spatial and Temporal Adaptive Reflectance Fusion Model) algorithm to multi-spectral images covering two semi-arid heterogeneous rangeland study sites located in South Africa. MODIS 250 m resolution and RapidEye 5 m resolution images were fused to produce synthetic RapidEye images, from June 2011 to July 2012. We evaluated the performance of the algorithm by comparing predicted surface reflectance values to real RapidEye images. Our results show that ESTARFM predictions are accurate, with a coefficient of determination for the red band 0.80 < R2 < 0.92, and for the near-infrared band 0.83 < R2 < 0.93, a mean relative bias between 6% and 12% for the red band and 4% to 9% in the near-infrared band. Heterogeneous vegetation at sub-MODIS resolution is captured adequately: A comparison of NDVI time series derived from RapidEye and ESTARFM data shows that the characteristic phenological dynamics of different vegetation types are reproduced well. We conclude that the ESTARFM algorithm allows us to produce synthetic remote sensing images at high spatial combined with high temporal resolution and so provides valuable information on vegetation dynamics in semi-arid, heterogeneous rangeland landscapes.
Highlights
Time series of vegetation indices derived from remotely sensed data are readily used to monitor spatial and temporal dynamics of biophysical variables such as Leaf Area Index (LAI) or fraction of absorbed photosynthetically active radiation [1,2], and phenological metrics such as day of green-up and duration of greenness [3]
The ESTARFM algorithm could successfully be applied for the generation of a high spatio-temporal vegetation time series by fusing RapidEye and MODIS imagery
The ESTARFM prediction accuracy is good for the red and NIR bands during phases of little vegetation dynamics, but deteriorates during times of quick vegetation growth
Summary
Time series of vegetation indices derived from remotely sensed data are readily used to monitor spatial and temporal dynamics of biophysical variables such as Leaf Area Index (LAI) or fraction of absorbed photosynthetically active radiation (fAPAR) [1,2], and phenological metrics such as day of green-up and duration of greenness [3]. Multi-temporal measurements of those variables provide information for spatially distributed environmental modelling across different scales, e.g., for yield forecasts, ecohydrologic cycles or land cover change [4]. Ecological studies often rely solely on field data, whose collection is time-consuming, cost-intensive and often limited in space and time. Relevant processes may display a small-scale heterogeneity (e.g., spatially varying grazing intensities or heterogeneous abiotic site conditions [6]) that a selective sampling of species may not capture
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
