Remote sensing has many applications in forestry. Research has been carried out to ascertain the potential of estimating forest volume from remotely sensed imagery (RSI), but less attention has concentrated on the estimation of forest biomass. Research that has considered volume and/or biomass assessment has encountered problems regarding saturation of image features, low sampling quantity/extent and poor methodologies for biomass calculation. The increasing usage of biomass as a fuelcrop combined with wider concerns of carbon sequestration and the effect of whole tree harvesting regimes on soil nutrient status provide a real stimulus to solve these problems and to extend research to include biomass.
This research used LANDSAT Thematic Mapper (TM) and Airborne Thematic Mapper (ATM) RSI to estimate volume (m3/ha) and above ground biomass (stem, branches, foliage) components (dry t/ha) within a Pinus nigra var maritima plantation forest in east Anglia, U.K. Estimates of biomass were collected in 1993 and 1987 for 60 stands aged from 1368 years old. Stand volumes ranged from 50500 m3/ha. Biomass estimates 1993/1989 were regressed against radiometrically calibrated and atmospherically corrected RSI data from the TM and ATM respectively. The results from the ATM data were particularly accurate, estimating total above ground biomass and volume to an accuracy of 22.4 dry t/ha and 45.9 m3/ha (R2 = 0.87, R2 = 0.89) respectively. Estimates of foliage biomass were less accurate (2.12 t/ha, R2 = 0.26). The results from the TM imagery show similar trends in prediction but with reduced accuracy.
The poster concludes by considering the validity in extending this technique of biomass and volume assessment to other species/regions and compares the accuracy and cost effectiveness of RSI with ground mensuration techniques. A range of RSI images are presented which show the potential of RSI in forest evaluation and monitoring.
Key words: remotely sensed imagery, forest biomass, yield prediction.