Wavelet-based detection of bush encroachment in a savanna using multi-temporal aerial photographs and satellite imagery |
| |
| Institution: | 1. CEFE, UMR 5175, CNRS, 1919 Route de Mende, 34293 Montpellier Cedex 5, France;2. Dynafor, UMR 1202, Université de Toulouse, INRA, INPT-ENSAT, INPT - EI Purpan, BP 52627, 31326 Castanet – Tolosan Cedex, France;3. CEFE, UMR 5175, Ecole Pratique des Hautes Etudes, Biogéographie et Ecologie des Vertébrés, campus CNRS, 1919 route de Mende, 34 293 Montpellier Cedex 5, France;4. Département Ecologie et Gestion de la Biodiversité, Muséum National d''Histoire Naturelle, 57 rue Cuvier, 75005 Paris Cedex 5, France;5. AGAP, UMR 1334, INRA, 2 place Pierre Viala, 34060 Montpellier Cedex 2, France;1. School of Sustainable Engineering and the Built Environment, Arizona State University, Tempe, AZ 85287, USA;2. School of Earth and Space Exploration, Arizona State University, Tempe, AZ 85287, USA;3. Departamento de Ingeniería Civil y Ambiental, Universidad Autónoma de Ciudad Juárez, Ciudad Juárez, México;4. USDA-ARS Southwest Watershed Research Center, 2000 E. Allen Road, Tucson, AZ 85719, USA;5. School of Natural Resources and the Environment, University of Arizona, Tucson, AZ 85721, USA;6. Department of Earth, Environmental and Geographical Sciences, Northern Michigan University, Marquette, MI 49855, USA |
| |
| Abstract: | Although increased woody plant abundance has been reported in tropical savannas worldwide, techniques for detecting the direction and magnitude of change are mostly based on visual interpretation of historical aerial photography or textural analysis of multi-temporal satellite images. These techniques are prone to human error and do not permit integration of remotely sensed data from diverse sources. Here, we integrate aerial photographs with high spatial resolution satellite imagery and use a discrete wavelet transform to objectively detect the dynamics in bush encroachment at two protected Zimbabwean savanna sites. Based on the recently introduced intensity-dominant scale approach, we test the hypotheses that: (1) the encroachment of woody patches into the surrounding grassland matrix causes a shift in the dominant scale. This shift in the dominant scale can be detected using a discrete wavelet transform regardless of whether aerial photography and satellite data are used; and (2) as the woody patch size stabilises, woody cover tends to increase thereby triggering changes in intensity. The results show that at the first site where tree patches were already established (Lake Chivero Game Reserve), between 1972 and 1984 the dominant scale of woody patches initially increased from 8 m before stabilising at 16 m and 32 m between 1984 and 2012 while the intensity fluctuated during the same period. In contrast, at the second site, which was formely grass-dominated site (Kyle Game Reserve), we observed an unclear dominant scale (1972) which later becomes distinct in 1985, 1996 and 2012. Over the same period, the intensity increased. Our results imply that using our approach we can detect and quantify woody/bush patch dynamics in savanna landscapes. |
| |
| Keywords: | Dominant scale Intensity Wavelet transform Bush encroachment Savanna Remote sensing |
| 本文献已被 ScienceDirect 等数据库收录！ |
|
