Lacina Coulibaly

Intégration de données d'images satellitaires optiques et radars et de données topographiques pour la cartographie géomorphologique

In this paper, we study the contribution of remote sensing data, topographic data, and geoscience ancillary data to surficial deposits mapping. The performance of SPOT-4, Landsat-7, and RADARSAT-1 SAR satellite data were compared to identify and map surficial deposits. Different spectral indices (normalized difference vegetation index (NDVI), transformed soil adjusted vegetation index (TSAVI), redness index (RI), form index (IF), coloration index (CI), intensity, hue, and saturation) and textural features (mean, standard deviation, angular second moment, entropy, etc.) were extracted from these datasets and used in the mapping process. Morphometric data extracted from the digital elevation model (DEM) (altitude, slope, slope orientation, slope curvature, and potential moisture index) were integrated with the preceding datasets. A discriminant analysis was conducted to extract the most significant parameters, which were then used in a four-step linear combination mathematical model to map surficial deposits. We achieved an overall classification rate of 70% using spectral data of land use map in step 1. By adding information on vegetation and soils obtained from evaluation of spectral indices, this rate was improved to 79% during step 2. The addition of texture parameters to the previous information increased the classification accuracy to 81%. Finally, the addition of topographical information to the datasets of the previous step provided a further improvement in the global classification rate that increased to 88%.

Mapping Aboveground Forest Biomass from Ikonos Satellite Image and Multi-Source Geospatial Data using Neural Networks and a Kriging Interpolation

The present study develops a method for aboveground forest biomass mapping from Ikonos imagery and geospatial data. Reference biomass values by group of species were estimated using Kers equations and inventory data from permanent sample plots (PEP) of 400 m2. A supervised classification of the Ikonos image, based on the maximum likelihood method presenting the five species groups inventoried in the field study, was carried out. Thereafter, various vegetation indices and texture parameters were extracted from the Ikonos image. The extracted Ikonos data were then combined with geospatial data at the same 1 m spatial resolution. Inventory plots biomass values estimated by group of species were used for the neural networks model (Multi-layer Perceptron) training with the backpropagation algorithm. Thereafter, biomass values for sample pixels generated randomly by group of species were predicted with the Multi-layer Perceptron. Then, sample pixels biomass values of each group were used to derive biomass values of other pixels of the same species group by interpolation with the ordinary kriging method using five different variogram models. The Gaussian variogram model yielded the best biomass estimates by comparison with reference biomass values, with percentages of residual errors ranging between 2,6 and 9,8% (absolute value) and percentages of RMSE (root mean square error) ranging between 17.2 and 61.1%.

Mapping forest stands using RADARSAT-2 quad-polarization SAR images: A combination of polarimetric and spatial information

This study proposes an approach which simultaneously uses spatial information and polarimetric data from a RADARSAT-2 quad-polarization satellite image for forest tree species classification. The study area is near the Gounamitz River located in northwestern New Brunswick (Canada). After geometric correction of the image, two statistical models were used for the classification: (1) a Markov random fields model based on an initial segmentation provided by the K-means algorithm to account for the spatial statistical dependencies between adjacent sites; and (2) a K-distribution model with, as parameters, the covariance matrix containing all of the polarimetric information. The classification was optimized using the stochastic simulated annealing algorithm. Validation of the results was performed by comparison with field inventory measurements. Variation of the backscattering coefficient c° obtained for the RADARSAT-2 quad-polarization SAR image with incidence angles of 26 0 and 45 ° ranged from 1 and 3 dB for the different tree species. The results of average and overall accuracies of the classification were respectively 77.13% and 72.35% for the 26° incidence angle image compared to 81.47% and 79.12% for the 45°incidence angle.

Estimating forest stands vigor from airborne images and neural networks

The vigor of a tree defines its ability to grow and is associated with its productivity ([1]; [2]). The measurement of a trees vigor makes it possible to evaluate its general state of health and its development over time. This task is traditionally performed by identifying defects affecting the vigor such as visible signs and symptoms on the tree, like fungi on the main stem, significant forks, cracks, wounds, competition and the percentage of living crown ([2]). The vigor and quality of trees are important factors used by forest managers in the application of silvicultural prescriptions at stand and landscape levels. The estimation of the level of risk of vigor loss in forest stands thus becomes a major and essential issue for the management and the improvement of the productivity of forests in a context of sustainable development.

Combinaison de l'information spatiale et polarimétrique pour la cartographie de la composition forestière à partir d'images ROS RADARSAT-2

The present study presents an approach simultaneously using the spatial information and polarimetric information provided by RADARSAT-2 Quad Pol multipolarization image data for the classification of forest species. Two statistical models were used for classification purposes: (i) a Markov model taking into account the spatial statistical dependencies between adjacent sites based on an initial segmentation derived from the K-means algorithm and, (ii) a K distribution model using as parameters the covariance matrix containing all the polarimetric information and the shape parameter characteristic of the K distribution estimated using the moments of the theoretical and practical normalized intensities. The classification is optimized using the stochastic simulated annealing (SA) algorithm. Validation of the results was carried out through comparison with ground data observations. The variation of the backscattering coefficient σ° obtained for the RADARSAT-2 Quad Pol multipolarization images with incidence angles of 26° and 45° is equal to 3 dB for the different types of tree species stands. Using HH, VV, and HV linear polarizations it was possible to discriminate only four classes (watercourses, tolerant hardwoods, intolerant hardwoods, and conifers), with only a slight interclass difference of 1 dB. With a modification of the incidence angle from 26° to 45°, no significant change in the variation of the backscattering coefficient was noted in relation to the different types of tree species. The mean and overall precision results obtained for the classification are 81.47% and 79.12%, respectively, for the image with a 45° incidence angle and 77.13% and 72.35% for the image with a 26° incidence angle. [Traduit par la Rédaction]

Development of a cartographic index of forest stand vitality using an ikonos satellite image

This study develops a cartographic index of forest stands vitality (vigour) from Ikonos imagery data. Dendrometric data (dbh, height, age) were collected during field inventories of various forest stands mainly composed of fir, intolerant hardwoods and tolerant hardwoods species typical of the study area of Gounamitz, in the north-west of New Brunswick. These data were used to compute reference estimates of stand vitality for each sample stand, using the Lebels vitality equation [1]. Subsequently, corrected reflectance values for inventoried tree species were extracted from an Ikonos satellite image of the study area, after fusion, segmentation and classification were performed. Vegetation (NDVI, MSAVI, TSAVI, etc.) and texture indices were also extracted from the Ikonos image. Regression models (linear, polynomial and logarithmic) were then established between reference estimates of stand vitality computed earlier and extracted remote sensing data. Three of multiple regressions were tested. Obtained results showed that the linear regression model yielded the best estimates of forest stands vitality. Determination coefficients (r 2) were 0.771, 0.783 and 0.776 respectively for fir, intolerant hardwoods and tolerant hardwoods species. Further, the regression model used was validated by comparing for 36 reference plots, the values of vitality computed from forest inventory data to values estimated by the model. The overall regression on the studied fir, tolerant and intolerant hardwood stands generated RMSE, of 0.17, 0.29 and 0.10, respectively. In a second step, the results of this study were generalized to map the vitality of forest stands throughout the whole area of the study.