Informations et ressources scientifiques
sur le développement des zones arides et semi-arides

Accueil du site → Master → Etats Unis → 2015 → Differentiating Anthropogenic Rangeland Degradation from Climate Variability on Cimarron National Grassland in Morton County, Kansas

California State University, Northridge (2015)

Differentiating Anthropogenic Rangeland Degradation from Climate Variability on Cimarron National Grassland in Morton County, Kansas

Van Buskirk, Joshua

Titre : Differentiating Anthropogenic Rangeland Degradation from Climate Variability on Cimarron National Grassland in Morton County, Kansas

Auteur : Van Buskirk, Joshua

Université de soutenance : California State University, Northridge

Grade : Master of Arts (MA) 2015

Présentation
Degradation of public grazing land by cattle on Cimarron National Grassland in Morton County, Kansas, USA between 1991 and 2000 was studied using Landsat remote sensing data. The normalized difference vegetation index was calculated from the satellite data and used as an indicator of relative density and health of vegetation. Distinguishing the change in vegetation health due to natural factors (weather) and grazing is difficult because plant growth in arid and semiarid environments is often highly dependent on rainfall. Therefore, the climate signal must be removed from vegetation data prior to assessing human-induced degradation. To assess the change in vegetation health due to natural factors (weather), a regression model was developed based on the relationship between greenness and precipitation, and deviations from this model were attributed to grazing. These deviations (residuals) were studied to determine whether a relationship existed between them and grazing data. Residuals from this model were compared to both the number of days that grazing took place and animal units summed over three different timeframes individually. Correlation tests revealed a significant negative correlation between NDVI residuals and the cumulative number of grazing days in many parts of the study area. Animal units were not significantly correlated with residuals. Future research into the best method for accumulating grazing data over different intervals is still needed. Results show that the residual trend method can effectively remove the climate signal from remotely sensed vegetation data.

Présentation

Version intégrale (1,20 Mb)

Page publiée le 4 novembre 2016, mise à jour le 12 septembre 2017