Pol Coppin

Digital change detection in forest ecosystems with remote sensing imagery

Abstract The worlds forest ecosystems are in a state of permanent flux at a variety of spatial and temporal scales. Monitoring techniques based on multispectral satellite‐acquired data have demonstrated potential as a means to detect, identify, and map changes in forest cover. This paper, which reviews the methods and the results of digital change detection primarily in temperate forest ecosystems, has two major components. First, the different perspectives from which the variability in the change event has been approached are summarized, and the appropriate choice of digital imagery acquisition dates and interval length for change detection are discussed. In the second part, preprocessing routines to establish a more direct linkage between digital remote sensing data and biophysical phenomena, and the actual change detection methods themselves are reviewed and critically assessed. A case study in temperate forests (north‐central U.S.A.) then serves as an illustration of how the different change detectio...

Processing of multitemporal Landsat TM imagery to optimize extraction of forest cover change features

Digital procedures to optimize the information content of multitemporal Landsat TM data sets for forest cover change detection are described. Imagery from three different years (1984, 1986, and 1990) were calibrated to exoatmospheric reflectance to minimize sensor calibration offsets and standardize data acquisition aspects. Geometric rectification was followed by atmospheric normalization and correction routines. The normalization consisted of a statistical regression over time based on spatially well-defined and spectrally stable landscape features spanning the entire reflectance range. Linear correlation coefficients for all bitemporal band pairs ranged from 0.9884 to 0.9998. The correction mechanism used a dark object subtraction technique incorporating published values of water reflectance. The association between digital data and forest cover was maximized and interpretability enhanced by converting band-specific reflectance values into vegetation indexes. Bitemporal vegetation index pairs for each time interval (two, four, and six years) were subjected to two change detection algorithms, standardized differencing and selective principal component analysis. Optimal feature selection was based on statistical divergence measures. Although limited to spectrally-radiometrically defined change classes, results show that the relationship between reflective TM data and forest canopy change is explicit enough to be of operational use in a forest cover change stratification phase prior to a more detailed assessment. >

Optimization of the Information Content of Multitemporal Landsat Tm Data Sets for Monitoring Forest Cover Disturbance

The objective of this paper is to describe procedures that were developed to optimize the information content of multitemporal TM data sets for forest cover disturbance monitoring in Minnesota To minimize sensor calibration offsets and standardize data acquisition aspects, TM imagery from three different years (1984, 1986 and 1990) were first calibrated to exoatmospheric reflectance. After geometric rectification, an atmospheric correction routine was applied combining two major components, atmospheric normalization over time and transformation to ground reflectance. Atmospheric conditions were modelled over time using regression functions derived from five ground features known to be unchanged over the time interval of interest and spanning the entire image reflectance range. Linear correlation coefficients for all bitemporal band pairs ranged from 0.9884 to 0.9998. The correlation between digital data and the ground feature of interest, the forest cover, was subsequently maximized and irrelevant information content was reduced by converting the band-specific reflectances into seven vegetation indices that were assumed to cany unique information. The application of two change detection algorithms to these seven indices ultimately resulted in 14 change features for each time interval of interest (two, four and six years). Results have shown that the preprocessing sequence is absolutely vital to the forest cover monitoring methodology. The success rate for the detection of stand-based canopy depletion events over the six year interval was %%, while the pixel-based thematic accuracy assessment resulted in an average class accuracy of 91% with a KHAT value of 0.82.

Terrestrial laser-scanning provides means for the analytical evaluation of plantation systems

This paper presents a theoretical concept for the mechanistic analysis of plantation systems in terms of physiological efficiency that comprises three steps: 1. The automated evaluation of terrestrial 3D-laser scanner data of apple trees, 2. Tree reconstruction based on allometric relationships between branch diameter and leaf area, and 3. A fine-scaled 3D-light- and photosynthesis model. The automated evaluation procedure for laser-scanner data based on the 3D-Hough transformation could correctly identify the point-to-point connections of the main branch system of apple trees. Based on 3D-structure measurements, the 3D-light model STANDFLUX-SECTORS predicted maximum leaf mass per area (LMAmax) of branches with a root mean square error of 10.3 g/m2. An application of the model to beech trees revealed parallel courses of modelled light climate and measured sapflow of branches.

CLabSpeG: a compact laboratory spectro-goniometer system enabling rapid and complete BRDF assessments of forest elements

This paper presents a new multispectral compact laboratory gonioreflectometer that can adequately sample bidirectional reflectance functions. It will be used for obtaining BRDF data of vegetation elements that can be applied in light transport simulations within virtual forests. In this way, future investigation and interpretation of radiation of forest canopies will be made possible.