COMBINING PHOTOGRAMMETRIC CAMERA AND IR VIDEOGRAPHY TO DEFINE WITHIN-FIELD SOIL SAMPLING FRAMEWORKS

This paper investigates the use of photogrammetric camera and IR videography data to improve the design of field survey sampling frameworks. Spatial data collection can contribute up to 80% of the cost of deploying a GIS (Geographic Information System) based land use decision support system. The use of remotely sensed information and geostatistical methods combined with field survey using dGPS (differential Global Positioning System) is expected to maximise data quality while minimising costs. The remotely sensed data used were medium format colour photography and IR (Infra-Red) videography. These were orthorectified to the national (Ordnance Survey) map base and mosaiced using ERDAS Imagine. ERDAS Imagine was also used to combine the blue, green and red layers of the colour photography with the IR videography into a single four layer image. The stratified sampling strategy adopted to date allocates four points per hectare randomly within individual field boundaries. The sample points, generated within ArcView, are located using dGPS. The data are then interpolated to a grid using geostatistics. A second strategy uses the remotely sensed information to identify within-field variability by means of classified soil or NDVI models. The sample sites are sub-stratified (at 4 per hectare) by variability classes with a minimum mapping unit of 0.25 hectare. Both strategies were employed at a test site and the results evaluated against validation samples collected on a 100m grid. The use of the remotely sensed information ensured that the survey sampled the range of within-field variability.