Remote-sensing monitoring of desertification, phenology, and droughts

Arnon Karnieli, Giorgio Dall'Olmo

    Research output: Contribution to journalReview articlepeer-review

    28 Scopus citations


    Year-to-year fluctuations of rainfall in the northern Negev desert provide an opportunity to characterize and assess the temporal dynamics of desertification, phenology, and drought processes. Such information was retrieved and analyzed by combined use of satellite imageries in the reflectivity and thermal spectral bands. Data covering four years of coarse spatial resolution and images from a high revisit time satellite, namely the NOAA-14, were used. The images were processed to produce the normalized difference vegetation index (NDVI) and the land surface temperature (LST). These measures were applied to the sand field in the northwestern Negev (Israel), which is almost totally covered by biological soil crusts, and to an adjacent region in Sinai (Egypt), consisting mainly of bare dune sands. Various manipulations of the data were applied. Time series presentation of the NDVI and LST reveals that the NDVI values correspond to the reaction of the vegetation to rainfall and that LST values represent seasonal climatic fluctuation. Scatterplot analysis of LST vs NDVI demonstrates the following: the two different biomes (Sinai and the Negev) exhibit different yearly variation of the phenological patterns (two seasons in Sinai moving along the LST axis, and three seasons in the Negev, where the NDVI axis represents the growing season); the Sinai has an ecosystem similar to that found in the Sahara, while the Negev, only a few kilometers away, has an ecosystem similar to the one found in the Sahel; and drought indicators were derived by using several geometrical expressions based on the two extreme points of the LST-NDVI scatterplot. The later analysis led to a discrimination function that aims to distinguish between the drought years and the wet years in both biomes. Results from the current study show that a great deal of information on dryland ecosystems can be derived from four, out of five, NOAA/AVHRR spectral bands. The NDVI is derived from the red and the near-infrared bands and the LST from the two thermal bands. Combined use of these two products provides more information than any product alone.

    Original languageEnglish
    Pages (from-to)22-38
    Number of pages17
    JournalManagement of Environmental Quality
    Issue number1
    StatePublished - 1 Mar 2003


    • Egypt
    • Environment
    • Geographical information systems
    • Israel

    ASJC Scopus subject areas

    • Public Health, Environmental and Occupational Health
    • Management, Monitoring, Policy and Law


    Dive into the research topics of 'Remote-sensing monitoring of desertification, phenology, and droughts'. Together they form a unique fingerprint.

    Cite this