Oumer S. Ahmed

Interpretation of forest disturbance using a time series of Landsat imagery and canopy structure from airborne lidar

In this study we examined forest disturbance, largely via forest harvest, over three decades in a coastal temperate forest on Vancouver Island, British Columbia, Canada. We analysed how disturbance history relates to current canopy structural conditions by interpreting the relationship between light detection and ranging (lidar) derived canopy structure and forest disturbance trajectories derived from Landsat images to assess if a particular stand structural condition is to result based on disturbance histories. The lidar data were obtained in 2004, and are used to relate forest structural conditions at the end of the Landsat time series (1972–2004), essentially providing for a measure of resultant structure emerging from the spectral trends captured. Correlation analysis was applied between lidar-derived canopy structure (canopy cover and height) and Landsat spectral indices, such as the Tasseled Cap Angle (TCA), which showed a strong correlation coefficient (r = 0.86) with canopy cover. TCA was then used to characterize change in forest disturbance through the full temporal depth of the available Landsat image time series using a trajectory-based characterization method. Approximately 71.5% of the study area was found to correspond to “stable and undisturbed forest”. Four disturbance classes (areas characterized by disturbance, disturbance followed by revegetation, ongoing revegetation, and revegetation to stable state) accounted for approximately 10.2%, 5.3%, 2.2%, and 10.5% of the study area, respectively. We evaluated the forest structural and spectral separability between the disturbance classes. In terms of structural variability the mean airborne lidar-derived canopy cover showed clear differentiation between disturbance classes. Spectral mixture analysis (SMA) was used to extract the spectral characteristics for each disturbance class. The SMA-derived fractions were then used to analyse the class separability between the Landsat trajectory derived disturbance classes. The fraction images provided clear distinction between disturbance classes in abundances between sunlit canopy, non-photosynthetic vegetation, shade, and exposed soil. The extracted spectral indices and SMA fractions within the Landsat trajectory derived disturbance classes were used to assess if terminal forest structural conditions can be related to a complex suite of stand development trajectories and processes. The Landsat spectral indices and SMA fractions were separately modeled to estimate lidar-derived mean canopy cover and height data within each disturbance class using multiple regression. The results indicate canopy cover and height regression models developed using spectral indices provided a relatively better estimation than those using SMA endmember fractions. Compared with the relatively regular structure of fully grown undisturbed (stable) forests, the forest disturbance classes typically exhibited complex irregular structure, making it more difficult to accurately estimate their canopy cover and height. As a result, all models developed for the stable forest class performed better than those developed for other forest disturbance classes. Modeling canopy cover and height from Landsat temporal spectral indices resulted in modeled agreement to lidar measures of R2 0.82 (RMSE 0.09) and R2 0.67 (RMSE 3.21), respectively. Our results also indicate moderately accurate predictions of lidar-derived canopy height can be obtained using the Landsat-level disturbance class endmember fractions with R2 0.60 and RMSE 4.19. This study demonstrates the potential of using the over four decade record of Landsat observations (since 1972) to estimate forest canopy cover and height using prestratification of the data based on disturbance trajectories.

Large Area Mapping of Annual Land Cover Dynamics Using Multitemporal Change Detection and Classification of Landsat Time Series Data

Abstract. Land cover characteristics remain of particular interest to the monitoring and reporting communities, and approaches for generating annual maps of land cover informed by change information derived from long time series are critically needed. In this study, we demonstrate and verify the utility of disturbance and recovery metrics derived from annual Landsat time series to inform the classification of annual land cover over a > 1.2 million hectare forest management area in the Boreal Mixedwood Region of northern Ontario, Canada. Annual land cover maps were generated, producing temporally informed products and compared to the established approach of using single-date spectral variables and indices. The Random Forest (RF) classification algorithm was used to classify land cover annually between 1990 and 2010, followed by the application of an annual temporal filter to remove illogical land cover transitions. Change detection in the study area had an overall accuracy of 92.47%. The use of time series metrics in the classification of land cover improved overall accuracy by 6.38% compared to single-date results. Using a separate independent reference sample, the RF classification approach combined with postclassification transition filtering resulted in an overall classification accuracy of 87.98%. The use of annual change and trend information to guide land cover, which is further informed by logical land cover transition rules, points to the creation of efficient, robust, and reliable land cover products in a transparent and operational fashion.

Estimation of Airborne Lidar-Derived Tropical Forest Canopy Height Using Landsat Time Series in Cambodia

In this study, we test and demonstrate the utility of disturbance and recovery information derived from annual Landsat time series to predict current forest vertical structure (as compared to the more common approaches, that consider a sample of airborne Lidar and single-date Landsat derived variables). Mean Canopy Height (MCH) was estimated separately using single date, time series, and the combination of single date and time series variables in multiple regression and random forest (RF) models. The combination of single date and time series variables, which integrate disturbance history over the entire time series, overall provided better MCH prediction than using either of the two sets of variables separately. In general, the RF models resulted in improved performance in all estimates over those using multiple regression. The lowest validation error was obtained using Landsat time series variables in a RF model (R2 = 0.75 and RMSE = 2.81 m). Combining single date and time series data was more effective when the RF model was used (opposed to multiple regression). The RMSE for RF mean canopy height prediction was reduced by 13.5% when combining the two sets of variables as compared to the 3.6% RMSE decline presented by multiple regression. This study demonstrates the value of airborne Lidar and long term Landsat observations to generate estimates of forest canopy height using the random forest algorithm.

Integration of Lidar and Landsat Data to Estimate Forest Canopy Cover in Coastal British Columbia

Disclaimer: The PDF document is a copy of the final version of this manuscript that was subsequently accepted by the journal for publication. The paper has been through peer review, but it has not been subject to any additional copy-editing or journal specific formatting (so will look different from the final version of record, which may be accessed following the DOI above depending on your access situation). Abstract Airborne Light Detection and Ranging (LiDAR) data provide useful measurements of forest canopy structure but are often limited in spatial coverage. Satellite remote sensing data from Landsat can provide extensive spatial coverage of generalized forest information. A forest survey approach that integrates airborne LiDAR and satellite data would potentially capitalize upon these distinctive characteristics. In this study in coastal forests of British Columbia, the main objective was to determine the potential of Landsat imagery to accurately estimate forest canopy cover measured from small-footprint airborne LiDAR data in order to expand the LiDAR measurements to a larger area. Landsat-derived Tasseled Cap Angle (TCA) and spectral mixture analysis (SMA) endmember fractions (i.e. sunlit canopy, non-photosynthetic vegetation (NPV), shade and exposed soil) were compared to LiDAR-derived canopy cover estimates. Pixel-based analysis and object-based area-weighted error calculations were used to assess regression model performance. The best canopy cover estimate was obtained (in the object-based deciduous forest models) with a mean object size (MOS) of 2.5 hectares (adjusted R 2 = 0.86 and RMSE = 0.28). Overall, lower canopy cover estimation accuracy was obtained for coniferous forests compared to deciduous forests in both the pixel and object-based approaches.

Classification of annual non-stand replacing boreal forest change in Canada using Landsat time series: a case study in northern Ontario

ABSTRACT Standardized protocols for forest change detection and classification based on Landsat time series data are becoming more common for use in characterizing multi-decadal history or trends in forest dynamics. One such protocol, referred to as Composite-2-Change (C2C), is a highly automated process developed in Canada that is applicable across extensive forest regions and includes change detection and typing based on Best-Available-Pixel image compositing, spectral trend analysis of breakpoints, object-based segmentation, and Random Forest classification. The aim of this article is to assess the classification accuracy of the Composite-2-Change (C2C) protocol in an eastern Canadian boreal forest environment in northern Ontario. Results demonstrated that the Landsat-derived change detection and attribution approach was approximately 90% accurate for stand-replacing forest change (fire, harvesting, roads), and approximately 75% for four non-stand replacing forest changes caused by spruce budworm and forest tent caterpillar defoliation, wetland and forest flooding caused by localized hydrological variations, and one class of multiple/other disturbances. The C2C protocol approach offers unique independent data layers for modelling that can be used to relate and inform on a range of substantive and subtle changes, which in turn can be labelled and tracked, offering otherwise unavailable information on forest dynamics over large areas.

The effects of topographic correction and gap filling in imagery on the detection of tropical forest disturbances using a Landsat time series in Myanmar

ABSTRACT In this study, we evaluated the effects of topographic correction and gap filling of Landsat Enhanced Thematic Mapper Plus (ETM+) images on the accuracy of forest change detection through a trajectory-based approach. Four types of Landsat time series stacks (LTSS) were generated. These stacks resulted from combinations of topographically corrected and uncorrected imagery combined with gap-filled and unfilled stacks. These combinations of stacks were then used as input into a trajectory-based change detection. The results of change detection from trajectory-based analysis using these LTSS were compared in order to assess the effects of both topographic correction and gap-filling procedures on the ability to detect forest disturbances. The results showed that overall accuracies of change detection were improved after gap filling (10.5% and 7.5%), but were only slightly improved after topographic correction (3.6% and 0.6%). Although the gap-filling process introduced some uncertainty that might have caused false change detection, the number of pixels whose detection of disturbance was enhanced after gap filling exceeded those detecting false change. The results also showed that the topographic correction did not contribute much to improve the change detection in this study area. However, topographic correction has a potential to increase the accuracy of change detection in areas of more rugged terrain and steep slopes. This is because a direct relationship between the slope of the topography with topographic correction and an enhanced detection of disturbance in pixels from year to year was observed in this study. For robust change detection, we recommend that a gap-filling process should be included in the trajectory-based analysis procedures such as the one used in this study where a single image per year is used to characterize change. We also recommend that in areas of rugged terrain, a topographic correction in the image pre-processing should be implemented.

Extending Airborne Lidar-Derived Estimates of Forest Canopy Cover and Height Over Large Areas Using kNN With Landsat Time Series Data

Airborne light detection and ranging (lidar) data provide an accurate and consistent means to obtain reliable forest canopy cover (CC) and height measurements, which are important in determining forest stand structure, volume, and biomass. Extending CC and height measurements over larger areas by integration with satellite imagery increases the value of airborne lidar data. A typical approach has been to use multiple regression, machine-learning, or regression tree methods to determine relationships between the forest structure variables measured in the lidar data and available single-date or multitemporal Landsat sensor reflectance data, and if the relationships are strong enough, to extend those variables over much larger areas than is typically covered by lidar. Such methods can be difficult to apply because of the complexity of the extending models and algorithms, and the long processing times, which may become prohibitive. One machine-learning approach, which uses the k-nearest neighbor (kNN) algorithm, is presented here in a British Columbia, Canada forest environment. Our goal was to simplify the estimation of lidar-derived forest structure variables with available Landsat time series data and compare the results of the kNN model to the traditional multiple regression results, and to those obtained with more complex and computationally demanding random forest (RF) methods. We develop and test the kNN model with airborne lidar-derived estimates of forest CC and height in 1846 relatively young and mature forests. The best kNN model produced estimates of airborne lidar-derived CC in validation sample (n = 1132) of mature forest with three Landsat time series variables with an R2 = 0.74 and RMSE of approximately 10%. This result is comparable to the results obtained in earlier work using more complex machine-learning approaches (in approximately 1/10 the time). Younger (i.e., recently disturbed) forest CC estimation was less successful in the kNN model because of the high degree of structural variability in these forests. Extending airborne lidar estimates of forest height to larger areas was also possible though, as expected, less successful than CC estimation using the kNN approach.

Wetland classification using Radarsat-2 SAR quad-polarization and Landsat-8 OLI spectral response data: a case study in the Hudson Bay Lowlands Ecoregion

ABSTRACT Circumboreal Canadian bogs and fens distinguished by differences in soils, hydrology, vegetation and morphological features were classified using combinations of Radarsat-2 synthetic aperture radar (SAR) quad-polarization data and Landsat-8 Operational Land Imager (OLI) spectral response patterns. Separate classifications were conducted using a traditional pixel-based maximum likelihood classifer and a machine learning algorithm following an object-based image analysis (OBIA). This study focused on two wetland classes with extensive coverage in the area (bog and fen). In the pixel-based maximum likelihood classification, accuracy increased from approximately 69% user’s accuracy and 79% producer’s accuracy using Radarsat-2 SAR data alone to approximately 80% user’s accuracy and 87% producer’s accuracy using Landsat-8 OLI data alone. Use of the Radarsat-2 SAR and Landsat-8 OLI data following principal components analysis (PCA) data fusion did not result in higher pixel-based maximum likelihood classification accuracy. In the object-based machine learning classification, higher bog and fen class accuracies were obtained when using Radarsat-2 and Landsat OLI data individually compared to the equivalent pixel-based classification. Subsequently, a PCA-data fusion product outperformed the individual bands of the Radarsat-2 and Landsat-8 imagery in object-based classification. Greater than 90% producer’s accuracy was obtained. The margin of error (MOE) was less than 5% in all classifications reported here. Further research will examine alternative data fusion techniques and the addition of Radarsat-2 SAR interferometric digital elevation model (DEM)-based geomorphometrics in object-based classification of different morphological types of bogs and fens.