Removal of Topographical Effect from Remote Sensing Data for Thermal Inertia Modelling

Topographical effect was considered in the context of bare soil moisture assessment. Soil moisture was evaluated using thermal inertia modelling. As an input data in thermal inertia model, the images in visible and thermal infrared range are required. Two own programs were worked out modi for thermal inertia modelling and topo – for topographical effect removal. Standard GIS software can have an option for topographical effect removal but only using two most popular (but not always correct) methods: Lambert’s and Minnaert’s. Theoretical test was performed to qualify significance of the topographical effect in thermal inertia modelling. For soil moisture evaluation on the base of the thermal inertia value, the correction of thermal image seems to be more important than correction of the image in visible range. To study the problem the remote sensing experiment was performed 27.10.1996 for test area (4x4 km) placed near Cracow in Poland. The satellite (Landsat) and airborne images were collected simultaneously with in situ measurements of soil moisture and temperature. Experiment was performed near meteorological station where standard meteorological parameters are measured. This kind of data was used for matching thermal inertia model. Conducted investigations confirm that topographical effects can significant influence remote sensing data. Proposed correction bases on assumption that the soils of the same moisture should “look” the same independent of slope and exposition. Topographical effect was easy to observe on the images in visible and infrared range on satellite and on airborne level. Different correction methods were tested for removal topographical effect from remote sensing data. After correction, correlation coefficient between DN and local illumination angle of the surface decreases from approximately 0.50 to 0.04. Initial results of the investigation connected to the influence of topographical effect on remote sensing data used in thermal inertia modelling were presented on the ISPRS Congress in Vienna in 1996. In the paper results of the following researches are presented.