Valerie A. Thomas

Image classification of a northern peatland complex using spectral and plant community data

Abstract Ordination and cluster analysis are two common methods used by plant ecologists to organize species abundance data into discrete “associations”. When applied together, they offer useful information about the relationships among species and the ecological processes occurring within a community. Remote sensing provides surrogate data for characterizing the spatial distribution of ecological classes based on the assumption of characteristic reflectance of species and species associations. Currently, there exists a need to establish and clarify the link between theories and practices of classification by ecologists and remote sensing scientists. In this study, high spatial resolution Compact Airborne Spectrographic Imager (CASI) reflectance data were examined and compared to plant community data for a peatland complex in northern Manitoba, Canada. The goal of this research was to explore the relationship between classification of species cover and community data and reflectance values. Ordination and cluster analysis techniques were used in conjunction with spectral separability measures to organize clusters of community-based data that were suitable for classification of CASI reflectance data, while still maintaining their ecological significance. Results demonstrated that two-way indicator species analysis (TWINSPAN) clusters did not correspond well to spectral reflectance and gave the lowest classification results of the methods investigated. The highest classification accuracies were achieved with ecological classes defined by combining the information obtained from a suite of analysis techniques (i.e., TWINSPAN, correspondence analysis (CA), and signature separability analysis), albeit not statistically superior to the classification obtained from the signature separability analysis alone.

On-the-Fly Massively Multitemporal Change Detection Using Statistical Quality Control Charts and Landsat Data

One challenge to implementing spectral change detection algorithms using multitemporal Landsat data is that key dates and periods are often missing from the record due to weather disturbances and lapses in continuous coverage. This paper presents a method that utilizes residuals from harmonic regression over years of Landsat data, in conjunction with statistical quality control charts, to signal subtle disturbances in vegetative cover. These charts are able to detect changes from both deforestation and subtler forest degradation and thinning. First, harmonic regression residuals are computed after fitting models to interannual training data. These residual time series are then subjected to Shewhart X-bar control charts and exponentially weighted moving average charts. The Shewhart X-bar charts are also utilized in the algorithm to generate a data-driven cloud filter, effectively removing clouds and cloud shadows on a location-specific basis. Disturbed pixels are indicated when the charts signal a deviation from data-driven control limits. The methods are applied to a collection of loblolly pine ( Pinus taeda) stands in Alabama, USA. The results are compared with stands for which known thinning has occurred at known times. The method yielded an overall accuracy of 85%, with the particular result that it provided afforestation/deforestation maps on a per-image basis, producing new maps with each successive incorporated image. These maps matched very well with observed changes in aerial photography over the test period. Accordingly, the method is highly recommended for on-the-fly change detection, for changes in both land use and land management within a given land use.

Fitting the Multitemporal Curve: A Fourier Series Approach to the Missing Data Problem in Remote Sensing Analysis

With the advent of free Landsat data stretching back decades, there has been a surge of interest in utilizing remotely sensed data in multitemporal analysis for estimation of biophysical parameters. Such analysis is confounded by cloud cover and other image-specific problems, which result in missing data at various aperiodic times of the year. While there is a wealth of information contained in remotely sensed time series, the analysis of such time series is severely limited due to the missing data. This paper illustrates a technique which can greatly expand the possibilities of such analyses, a Fourier regression algorithm, here on time series of normalized difference vegetation indices (NDVIs) for Landsat pixels with 30-m resolution. It compares the results with those using the spatial and temporal adaptive reflectance fusion model (STAR-FM), a popular approach that depends on having MODIS pixels with resolutions of 250 m or coarser. STAR-FM uses changes in the MODIS pixels as a template for predicting changes in the Landsat pixels. Fourier regression had an R2 of at least 90% over three quarters of all pixels, and it had the highest RPredicted2 values (compared to STAR-FM) on two thirds of the pixels. The typical root-mean-square error for Fourier regression fitting was about 0.05 for NDVI, ranging from 0 to 1. This indicates that Fourier regression may be used to interpolate missing data for multitemporal analysis at the Landsat scale, especially for annual or longer studies.

Canopy chlorophyll concentration estimation using hyperspectral and lidar data for a boreal mixedwood forest in northern Ontario, Canada

This study investigates the potential of lidar and hyperspectral data for prediction of canopy chlorophyll (Chl) and carotenoid concentrations for a spatially complex boreal mixedwood. First, canopy scale application of hyperspectral reflectance and derivative indices are used to estimate Chl concentration. Second, lidar data analyses is conducted to identify structural metrics related to Chl concentration. Third, lidar metrics and hyperspectral indices are combined to determine if Chl concentration estimates can be improved further. Of the hyperspectral indices considered, only the derivative chlorophyll index (DCI) and the red‐edge inflection point (λp) are shown to be good predictors of Chl concentration when mixed‐species plots are included in the analysis (i.e., for total chlorophyll concentration (a+b), r 2 = 0.79, RMSE = 4.6 µg cm−2 and r 2 = 0.78, RMSE = 4.5 µg cm−2 for DCI and λp, respectively). Integrating mean lidar first return heights for the 25th percentile with the hyperspectral DCI index further strengthens the relationship to canopy Chl concentration (i.e., for Chl(a+b), r 2 = 0.84, RMSE = 3.5 µg cm−2). Maps of total chlorophyll concentration for the study site reveal distinct spatial patterns that are indicative of the spatial distribution of species at the site.

Leaf area and clumping indices for a boreal mixed-wood forest: lidar, hyperspectral, and Landsat models

The leaf area index (LAI) and the clumping index (CI) provide valuable insight into the spatial patterns of forest canopies, the canopy light regime and forest productivity. This study examines the spatial patterns of LAI and CI in a boreal mixed-wood forest, using extensive field measurements and remote sensing analysis. The objectives of this study are to: (1) examine the utility of airborne lidar (light detection and ranging) and hyperspectral data to model LAI and clumping indices; (2) compare these results to those found from commonly used Landsat vegetation indices (i.e. the normalized difference vegetation index (NDVI) and the simple ratio (SR)); (3) determine whether the fusion of lidar data with Landsat and/or hyperspectral data will improve the ability to model clumping and LAI; and (4) assess the relationships between clumping, LAI and canopy biochemistry. Regression models to predict CI were much stronger than those for LAI at the site. Lidar was the single best predictor of CI (r 2 > 0.8). Landsat NDVI and SR also had a moderately strong predictive performance for CI (r 2 > 0.68 with simple linear and non-linear regression forms), suggesting that canopy clumping can be predicted operationally from satellite platforms, at least in boreal mixed-wood environments. Foliar biochemistry, specifically canopy chlorophyll, carotenoids, magnesium, phosphorus and nitrogen, was strongly related to the clumping index. Combined, these results suggest that Landsat models of clumping could provide insight into the spatial distribution of foliar biochemistry, and thereby photosynthetic capacity, for boreal mixed-wood canopies. LAI models were weak (r 2 < 0.4) unless separate models were used for deciduous and coniferous plots. Coniferous LAI was easier to model than deciduous LAI (r 2 > 0.8 for several indices). Deciduous models of LAI were weaker for all remote sensing indices (r 2 < 0.67). There was a strong, linear relationship between foliar biochemistry and LAI for the deciduous plots. Overall, our results suggest that broadband satellite indices have strong predictive performance for clumping, but that airborne hyperspectral or lidar data are required to develop strong models of LAI at this boreal mixed-wood site.

Investigating the Utility of Wavelet Transforms for Inverting a 3-D Radiative Transfer Model Using Hyperspectral Data to Retrieve Forest LAI

The need for an efficient and standard technique for optimal spectral sampling of hyperspectral data during the inversion of canopy reflectance models has been the subject of many studies. The objective of this study was to investigate the utility of the discrete wavelet transform (DWT) for extracting useful features from hyperspectral data with which forest LAI can be estimated through inversion of a three dimensional radiative transfer model, the Discrete Anisotropy Radiative Transfer (DART) model. DART, coupled with the leaf optical properties model PROSPECT, was inverted with AVIRIS data using a look-up-table (LUT)-based inversion approach. We used AVIRIS data and in situ LAI measurements from two different hardwood forested sites in Wisconsin, USA. Prior to inversion, model-simulated and AVIRIS hyperspectral data were transformed into discrete wavelet coefficients using Haar wavelets. The LUT inversion was performed with three different datasets, the original reflectance bands, the full set of wavelet extracted features, and two wavelet subsets containing 99.99% and 99.0% of the cumulative energy of the original signal. The energy subset containing 99.99% of the cumulative signal energy provided better estimates of LAI (RMSE = 0.46, R2 = 0.77) than the original spectral bands (RMSE = 0.60, R2 = 0.47). The results indicate that the discrete wavelet transform can increase the accuracy of LAI estimates by improving the LUT-based inversion of DART (and, potentially, by implication, other terrestrial radiative transfer models) using hyperspectral data. The improvement in accuracy of LAI estimates is potentially due to different properties of wavelet analysis such as multi-scale representation, dimensionality reduction, and noise removal.

Combined Use of Airborne Lidar and DBInSAR Data to Estimate LAI in Temperate Mixed Forests

The objective of this study was to determine whether leaf area index (LAI) in temperate mixed forests is best estimated using multiple-return airborne laser scanning (lidar) data or dual-band, single-pass interferometric synthetic aperture radar data (from GeoSAR) alone, or both in combination. In situ measurements of LAI were made using the LiCor LAI-2000 Plant Canopy Analyzer on 61 plots (21 hardwood, 36 pine, 4 mixed pine hardwood; stand age ranging from 12-164 years; mean height ranging from 0.4 to 41.2 m) in the Appomattox-Buckingham State Forest, Virginia, USA. Lidar distributional metrics were calculated for all returns and for ten one meter deep crown density slices (a new metric), five above and five below the mode of the vegetation returns for each plot. GeoSAR metrics were calculated from the X-band backscatter coefficients (four looks) as well as both X- and P-band interferometric heights and magnitudes for each plot. Lidar metrics alone explained 69% of the variability in LAI, while GeoSAR metrics alone explained 52%. However, combining the lidar and GeoSAR metrics increased the R2 to 0.77 with a CV-RMSE of 0.42. This study indicates the clear potential for X-band backscatter and interferometric height (both now available from spaceborne sensors), when combined with small-footprint lidar data, to improve LAI estimation in temperate mixed forests.

An LUT-Based Inversion of DART Model to Estimate Forest LAI from Hyperspectral Data

The efficient inversion of complex, three-dimensional (3-D) radiative transfer models (RTMs), such as the discrete anisotropy radiative transfer (DART) model, can be achieved using a look-up table (LUT) approach. A pressing research priority in LUT-based inversion for a 3-D model is to determine the optimal LUT grid size and density. We present a simple and computationally efficient approach for populating an LUT database with DART simulations over a large number of spectral bands. In the first step, we built a preliminary LUT using model parameters with coarse increments to simulate reflectance for six broad bands of Landsat Thematic Mapper (TM). In the second step, the preliminary LUT was compared with the TM reflectance, and the optimal input ranges and realistic parameter combinations that led to simulations close to Landsat spectra were then identified. In the third step, this information was combined with a sensitivity study, and final LUTs were built for the full spectrum of Airborne Visible/Infrared Imaging Spectrometer (AVIRIS) narrow bands and six Landsat broad bands. The final LUT was inverted to estimate leaf area index (LAI) in northern temperate forests from AVIRIS and TM data. The results indicate that the approach used in this study can be a useful strategy to estimate LAI accurately by DART model inversion.

Prediction of Macronutrients at the Canopy Level Using Spaceborne Imaging Spectroscopy and LiDAR Data in a Mixedwood Boreal Forest

Information on foliar macronutrients is required in order to understand plant physiological and ecosystem processes such as photosynthesis, nutrient cycling, respiration and cell wall formation. The ability to measure, model and map foliar macronutrients (nitrogen (N), phosphorus (P), potassium (K), calcium (Ca) and magnesium (Mg)) at the forest canopy level provides information on the spatial patterns of ecosystem processes (e.g., carbon exchange) and provides insight on forest condition and stress. Imaging spectroscopy (IS) has been used particularly for modeling N, using airborne and satellite imagery mostly in temperate and tropical forests. However, there has been very little research conducted at these scales to model P, K, Ca, and Mg and few studies have focused on boreal forests. We report results of a study of macronutrient modeling using spaceborne IS and airborne light detection and ranging (LiDAR) data for a mixedwood boreal forest canopy in northern Ontario, Canada. Models incorporating Hyperion data explained approximately 90% of the variation in canopy concentrations of N, P, and Mg; whereas the inclusion of LiDAR data significantly improved the prediction of canopy concentration of Ca (R2 = 0.80). The combined used of IS and LiDAR data significantly improved the prediction accuracy of canopy Ca and K concentration but decreased the prediction accuracy of canopy P concentration. The results indicate that the variability of macronutrient concentration due to interspecific and functional type differences at the site provides the basis for the relationship observed between the remote sensing measurements (i.e., IS and LiDAR) and macronutrient concentration. Crown closure and canopy height are the structural metrics that establish the connection between macronutrient concentration and IS and LiDAR data, respectively. The spatial distribution of macronutrient concentration at the canopy scale mimics functional type distribution at the site. The ability to predict canopy N, P, K, Ca and Mg in this study using only IS, only LiDAR or their combination demonstrates the excellent potential for mapping these macronutrients at canopy scales across larger geographic areas into the next decade with the launch of new IS satellite missions and by using spaceborne LiDAR data.

Testing the robustness of predictive models for chlorophyll generated from spaceborne imaging spectroscopy data for a mixedwood boreal forest canopy

The amount of chlorophyll in a leaf influences photosynthetic potential and can be an indicator of the overall condition of a plant, including its stress level and nutritional status. Hence, it is important to understand the spatial and temporal variation of chlorophyll concentration. Imaging spectroscopy (IS) has made it possible to estimate chlorophyll at leaf and canopy levels. Spaceborne imaging spectrometers offer the possibility of estimating chlorophyll concentration at larger spatial scales and at lower cost than from direct sampling. We undertook this study in a mixedwood boreal forest to test the robustness of predictive models generated using Hyperion data for predicting chlorophyll concentration of data sets from different locations collected in different years. Among the group of indices tested, the derivative chlorophyll index (DCI) (DCI = D705/D722) and the maximum derivative of the red-edge divided by the derivative of 703 nm (Dmax(680–750))/D703) emerged as the best predictors of chlorophyll concentration across space and through time. When the canopy level chlorophyll predictive models of DCI and Dmax(680–750))/D703 derived from Hyperion data were applied to other years’ remote-sensing data acquired by airborne and spaceborne sensors, DCI explained 71%, 63%, and 6% and Dmax(680–750))/D703 explained 61%, 54%, and 8% of the variation in chlorophyll in 2002, 2004, and 2008, respectively, with prediction errors ranging from 11.7% to 14.6%. Two-variable models generated using 2005 Hyperion data were not as robust for predicting chlorophyll concentration from other years. Two models were found to explain more than half of the variance in chlorophyll concentration for the 2004 data only. Single and two-variable models applied to 2008 chlorophyll data provided poor predictions. The presence of multiple species creates a gradient in the chlorophyll concentration, which makes it possible to predict chlorophyll concentration. The gradient also affects the performance of predictive models generated using data from a different year. However, differences in sensors may also affect model performance. Our results suggest that predictive models obtained from Hyperion data are robust in predicting chlorophyll concentration within the same site through time and also at different sites across sensors.