David L. Evans

Characterizing vertical forest structure using small-footprint airborne LiDAR

Abstract Characterization of forest attributes at fine scales is necessary to manage terrestrial resources in a manner that replicates, as closely as possible, natural ecological conditions. In forested ecosystems, management decisions are driven by variables such as forest composition, forest structure (both vertical and horizontal), and other ancillary data (i.e., topography, soils, slope, aspect, and disturbance regime dynamics). Vertical forest structure is difficult to quantify and yet is an important component in the decision-making process. This study investigated the use of light detection and ranging (LiDAR) data for classifying this attribute at landscape scales for inclusion into decision-support systems. Analysis of field-derived tree height variance demonstrated that this metric could distinguish between two classes of vertical forest structure. Analysis of LiDAR-derived tree height variance demonstrated that differences between single-story and multistory vertical structural classes could be detected. Landscape-scale classification of the two structure classes was 97% accurate. This study suggested that within forest types of the Intermountain West region of the United States, LiDAR-derived tree heights could be useful in the detection of differences in the continuous, nonthematic nature of vertical structure forest with acceptable accuracies.

Family influences on leaf area estimates derived from crown and tree dimensions in Pinus taeda

Abstract Numerous studies have attempted to use remote sensing tools to estimate leaf area index (LAI). However, these estimates have generally lacked precision and typically provide little information on stand or canopy structure. Recent advances in remote sensing technologies may allow more accurate estimates of LAI based upon measures of individual tree structure. In this paper, we describe how well leaf area (LA) of 15-year-old loblolly pine (Pinus taeda L.) trees can be estimated from tree and crown dimensional characteristics that might be derived using remote sensing tools. Because crown and canopy structure can vary genetically, we also tested for family differences in the allometric relationships. When family effects were not considered, the log of crown volume predicted log(LA) with an R2=0.65 compared to R2=0.81 for the relationship between LA and DBH. When tree height was included in the relationship with crown volume, R2 improved to 0.83. The log of crown width predicted log(LA) with an R2=0.58, and the log of crown length predicted log(LA) with an R2=0.44. Family effects were found to be significant in all of the models tested. With family effects accounted for, the log of crown volume predicted log(LA) with an R2=0.83 compared to 0.91 for the LA–DBH relationship. The combination of crown volume and tree height predicted log(LA) with an R2=0.91. The combination of tree height and height to middle of the crown predicted log(LA) with an R2=0.83 with family effects included. The log of crown width predicted log(LA) with an R2=0.80, and the log of tree height predicted log(LA) with an R2=0.75. Our analysis suggests that individual tree LA can be predicted as precisely from measures of tree and crown structural dimensions as it can from DBH. These relationships appear to be influenced by genetic factors, although no more so than DBH-based relationships. The most accurate LA estimates require information on both vertical and horizontal tree and crown dimensions. Given the current rate of technological advancement, it appears likely that remote sensing tools will be capable of providing accurate assessment of stand-level LAI in the near future.

LiDAR Forest Inventory with Single-Tree, Double-, and Single-Phase Procedures

Light Detection and Ranging (LiDAR) data at 0.5–2 m postings were used with double-sample, stratified procedures involving single-tree relationships in mixed, and single species stands to yield sampling errors ranging from ±2.1% to ±11.5%. LiDAR samples were selected with focal filter procedures and heights computed from interpolated canopy and DEM surfaces. Tree dbh and height data were obtained at various ratios of LiDAR, ground samples for DGPS located ground plots. Dbh-height and ground-LiDAR height models were used to predict dbh and compute Phase 2 estimates of basal area and volume. Phase 1 estimates were computed using the species probability distribution from ground plots in each strata. Phase 2 estimates were computed by randomly assigning LiDAR heights to species groups using a Monte Carlo simulation for each ground plot. There was no statistical difference between volume estimates from 0.5 m and 1 m LiDAR densities. Volume estimates from single-phase LiDAR procedures utilizing existing tree attributes and height bias relationships were obtained with sampling errors of 1.8% to 5.5%.

Multi-temporal analysis of landsat data to determine forest age classes for the mississippi statewide forest inventory~preliminary results

The use of Landsat data to aid in forest sampling stratification, area estimation, and future resource assessment through growth models is currently being investigated for the state of Mississippi with the goal of better understanding present and future wood resources. In such analyses, and as a part of this investigation, change detection techniques are being exploited to help determine these forest stand ages in approximate five year intervals. This preliminary report looks at post classification comparisons and temporal image differencing as two means to find these dates. The results find the post classification comparisons techniques, in an unrefined use, to work moderately well (overall accuracy = 0.6157, KHAT = 0.5386) and temporal image differencing with NDVI and tasseled cap transformations to disagree with each other in predicted age class sizes with no assessment data to validate accuracy at this time.

Scale-dependent den-site selection by American black bears in Mississippi

Abstract. Habitat selection is a hierarchical process in which animals select resources at varying spatial scales. Dens are a critical component of American black bear (Ursus americanus) habitat, yet scale-dependent den-site selection has received limited attention. Using habitat and topographic characteristics, we assessed scale-dependent den-site selection by 11 black bears during 20 den years in Mississippi, USA, at 3 spatial scales: den site (15 m), den area (100 m), and den landscape (1,000 m). Black bears in Mississippi exhibited scale-dependent den-site selection selecting greater percentage horizontal cover at the den area scale. Risk of flooding (i.e., elevation, distance to nearest stream), disturbance (i.e., distance to nearest stream or road), and habitat composition did not influence den-site selection at spatial scales measured. Greater percentage horizontal cover likely provides additional security and increases energetic efficiency. Selection for horizontal cover at finer spatial scales suggests security at den sites is a lower-order factor influencing fitness.

Predicting Southeastern Forest Canopy Heights and Fire Fuel Models using GLAS Data

The Geoscience Laser Altimeter System (GLAS) is a waveform lidar system carried on board the Ice, Cloud, and Elevation Satellite (ICESat). This study tested the use of GLAS data, from the L3e and L3g campaigns, to estimate total canopy height. GLAS footprint locations were sampled for reference data. The GLAS-derived and field-derived canopy heights portrayed good correlation (R 2 = 0.8354). This study also examined two representative fire fuel models within forests in East-Central Mississippi. GLAS waveforms were compared with field data for fire fuel models 9 and 10 of the fire fuel models described by Anderson (1982). GLAS data intensities were extracted and averaged to create predictive variables. Two variables were applied in Logistic regression to predict the probability of belonging to either fuel model (overall accuracy = 0.6875).

Change Detection Techniques for Use in a Statewide Forest Inventory Program

Eight change detection procedures and a hybrid forest type classification procedure were tested for their ability to detect forest land-cover change in east-central Mississippi. The best performing method was change vector analysis using vegetation indices with an image segmentation classification that produced an overall accuracy (82.50 percent) and overall Kappa (0.7900) calculated from error matrices. The hybrid forest type classification had an overall accuracy of 77.08 percent for 1997 and 71.25 percent for 2002. The results of this study were compared to a prior pilot inventory study for the same study area in east-central Mississippi. There was considerable disagreement between the two studies in terms of number of hectares in the age classes and the forest type classes, most likely attributed to the difference in methods for determining forest type classes.

Si Radial p-i-n Junction Photovoltaic Arrays with Built-In Light Concentrators.

High-performance photovoltaic (PV) devices require strong light absorption, low reflection and efficient photogenerated carrier collection for high quantum efficiency. Previous optical studies of vertical wires arrays have revealed that extremely efficient light absorption in the visible wavelengths is achievable. Photovoltaic studies have further advanced the wire approach by employing radial p-n junction architectures to achieve more efficient carrier collection. While radial p-n junction formation and optimized light absorption have independently been considered, PV efficiencies have further opportunities for enhancement by exploiting the radial p-n junction fabrication procedures to form arrays that simultaneously enhance both light absorption and carrier collection efficiency. Here we report a concept of morphology control to improve PV performance, light absorption and quantum efficiency of silicon radial p-i-n junction arrays. Surface energy minimization during vapor phase epitaxy is exploited to form match-head structures at the tips of the wires. The match-head structure acts as a built-in light concentrator and enhances optical absorptance and external quantum efficiencies by 30 to 40%, and PV efficiency under AM 1.5G illumination by 20% compared to cylindrical structures without match-heads. The design rules for these improvements with match-head arrays are systematically studied. This approach of process-enhanced control of three-dimensional Si morphologies provides a fab-compatible way to enhance the PV performance of Si radial p-n junction wire arrays.

Extracting digital terrain models in forestry using lidar data

Airborne light detection and ranging (LIDAR) is emerging as a tool to provide an accurate digital terrain model (DTM) of forest areas since it can even penetrate beneath the canopy. However, the determination of DTM in dense forest areas is still a difficult task and in an early stage of development. In this paper, an adaptive prediction technique based on the least mean squares (LMS) algorithm is presented. Results for LIDAR data, taken in 1999 at the Bellingham, WV site, are considered to illustrate the applicability of the presented technique.

Comparison of AVHRR classification and aerial photography interpretation for estimation of forest area

The USDA Forest Service Southern Forest Experiment Station`s Forest Inventory and Analysis (SO-FIA) unit uses a dot count method to estimate the percentage of forest area in counties or parishes from aerial photographs. The research reported in this paper was designed to determine whether Advanced Very High Resolution Radiometer (AVHRR) data could be used to estimate forest area at the county or parish level. For this study, AVHRR data for three parishes in central Louisiana were extracted from a 1991 AVHRR forest type map of the United States. Photo interpretation data were obtained from a digital mosaic of aerial photography of the parishes. Forest area estimates obtained by means of photo interpretation did not differ significantly from those obtained by analyzing AVHRR data.