Assessing variability in the production of pasture using GIS and remote sensing techniques

• Main Author: Smith, Steven Murray
• Language: English
• Published: University of British Columbia 2010
• Online Access: http://hdl.handle.net/2429/29293

Information relating to the spatial characteristics of biophysical resources has been difficult to incorporate into land management. In this study statistical analysis was used to demonstrate that forage yield and quality were influenced by the water balance and soil physical properties. Traditional empirical modelling techniques were of limited utility as predictors of yield and quality. However, multivariate statistical techniques provide predictor variables for individual forage cuts but not for a complete growing season. GIS provided several distinct advantages over traditional statistical techniques. First, it provided techniques to interpolate point data (such as forage yield and quality variables), and provide spatial distributions for a wide number of biophysical properties. Secondly, overlaying forage variables such as yield with a digital elevation model in a categoric manner provided output displaying the spatial relationships between the variables. Relationships derived from overlays using elevation and water retention properties provided good spatial predictions for several forage variables. Thirdly, digitized colour-IR aerial photographs were incorporated into the GIS where the pixel information was combined as map overlays via a regression equation. The resulting output provided very accurate spatial predictions for forage yield and quality parameters. Finally, economic data was generated in a spatial context and the resulting display was used to assess the effects of irrigation and management on forage yield and quality. The results suggest that the GIS techniques combined with remote sensing and economic data offer exciting possibilities to model and present spatial data. === Land and Food Systems, Faculty of === Graduate