Jessica J. Mitchell

Radiometric and Geometric Analysis of Hyperspectral Imagery Acquired from an Unmanned Aerial Vehicle

In the summer of 2010, an Unmanned Aerial Vehicle (UAV) hyperspectral calibration and characterization experiment of the Resonon PIKA II imaging spectrometer was conducted at the US Department of Energy’s Idaho National Laboratory (INL) UAV Research Park. The purpose of the experiment was to validate the radiometric calibration of the spectrometer and determine the georegistration accuracy achievable from the on-board global positioning system (GPS) and inertial navigation sensors (INS) under operational conditions. In order for low-cost hyperspectral systems to compete with larger systems flown on manned aircraft, they must be able to collect data suitable for quantitative scientific analysis. The results of the in-flight calibration experiment indicate an absolute average agreement of 96.3%, 93.7% and 85.7% for calibration tarps of 56%, 24%, and 2.5% reflectivity, respectively. The achieved planimetric accuracy was 4.6 m (based on RMSE) with a flying height of 344 m above ground level (AGL).

Subpixel abundance estimates in mixture-tuned matched filtering classifications of leafy spurge (Euphorbia esula L.)

Two demonstration sites in southeast Idaho, USA were used to extend remote sensing of leafy spurge research to fine-scale detection for abundance mapping using matched filtering (MF) scores. Linear regression analysis was used to quantify the relationship between MF estimates and calibrated ground estimates of leafy spurge abundance. The two sites had r 2 values of 0.46 and 0.64. Both the slope of the regressions and the scaling behaviour of MF scores indicate that the technique consistently underestimated true abundance (defined here as percentage canopy cover) by roughly one-third. This underestimation may be influenced by field estimation bias and algorithm confusion between target and background signal. Further results indicate that MF exhibits linear scaling behaviour in six locations containing dense, uniform infestations. At these locations, where canopy cover was held relatively constant, high spatial resolution (3 m) estimates were not significantly different from coarser spatial resolution estimates (up to 16 m). Given the mathematically unconstrained nature of the estimation technique, MF is not a straightforward method for estimating leafy spurge canopy cover.

Small-Footprint Lidar Estimations of Sagebrush Canopy Characteristics

The height and shape of shrub canopies are critical measurements for characterizing shrub steppe rangelands. Remote sensing technologies might provide an efficient method to acquire these measurements across large areas. This study compared point-cloud and rasterized lidar data to field-measured sagebrush height and shape to quantify the correlation between field-based and lidar-derived estimates. The results demonstrated that discrete return, small-footprint lidar with high point density (9.46 points/m 2 ) can provide strong predictions of true sagebrush height (R 2 of 0.84 to 0.86), but with a consistent underestimation of approximately 30 percent. Our results provided the first successful lidar-based descriptors of sagebrush shape with R 2 values of 0.65, 0.74, and 0.78 for respective predictions of shortest canopy diameter, longest canopy diameter, and canopy area. Future studies can extend lidar-derived shrub height and shape measurements to canopy volume, cover, and biomass estimates.

Vegetation and Slope Effects on Accuracy of a LiDAR-Derived DEM in the Sagebrush Steppe

This study analysed the errors associated with vegetation cover type and slope in a Light Detection and Ranging (LiDAR) derived digital elevation model (DEM) in a semiarid environment in southwest Idaho, USA. Reference data were collected over a range of vegetation cover types and slopes. Reference data were compared to bare-ground raster values and root mean square error (RMSE) and mean signed error (MSE) were used to quantify errors. Results indicate that vegetation cover type and slope have statistically significant effects on the accuracy of a LiDAR-derived bare-earth DEM. RMSE and MSE ranged from 0.072 to 0.220 m and from −0.154 to 0.017 m, respectively, with the largest errors associated with herbaceous cover and steep slopes. The lowest errors were associated with low sagebrush and low-slope environments. Although the RMSEs in this study were lower than those reported by others, further refinement of the accuracy of LiDAR systems may be needed for fine-scale vegetation and terrain applications in rangeland environments.

Leafy Spurge (Euphorbia esula) Classification Performance Using Hyperspectral and Multispectral Sensors

Abstract Two demonstration sites in southeast Idaho were used to extend the scope of remote sensing of leafy spurge research toward investigating coarser scale detection limits. Hyperspectral images were obtained to produce baseline leafy spurge maps, from which spatially and/or spectrally degraded images were subsequently derived for comparative purposes with Landsat 5 Thematic Mapper (TM). The baseline presence/absence maps had an overall accuracy of 67% at the Spencer study site and 85% at the Medicine Lodge study site. Unexpectedly high-accuracy results were produced from the images that were spectrally degraded to the bandwidths of Landsat 5 TM, which suggests that high spectral resolution is not critical to leafy spurge detection. However, a classification using a Landsat 5 TM image indicates that the sensor is inadequate for regional distribution monitoring. The differences in results between the actual and degraded images suggest that a sensor with comparable resolutions but improved instrumentation (e.g., signal to noise) may offer an alternative to hyperspectral data for mapping leafy spurge at regional scales.

Unmanned aerial vehicle (UAV) hyperspectral remote sensing for dryland vegetation monitoring

UAV-based hyperspectral remote sensing capabilities developed by the Idaho National Lab and Idaho State University, Boise Center Aerospace Lab, were recently tested via demonstration flights that explored the influence of altitude on geometric error, image mosaicking, and dryland vegetation classification. The test flights successfully acquired usable flightline data capable of supporting classifiable composite images. Unsupervised classification results support vegetation management objectives that rely on mapping shrub cover and distribution patterns. Overall, supervised classifications performed poorly despite spectral separability in the image-derived endmember pixels. In many cases, the supervised classifications accentuated noise or features in the mosaic that were artifacts of color balancing and “feathering” areas of flightline overlap. Future mapping efforts that leverage ground reference data, ultra-high spatial resolution photos and time series analysis should be able to effectively distinguish native grasses such as Sandberg bluegrass (Poa secunda), from invasives such as burr buttercup (Ranunculus testiculatus).

Single and Multi-Date Landsat Classifications of Basalt to Support Soil Survey Efforts

Basalt outcrops are significant features in the Western United States and consistently present challenges to Natural Resources Conservation Service (NRCS) soil mapping efforts. Current soil survey methods to estimate basalt outcrops involve field transects and are impractical for mapping regionally extensive areas. The purpose of this research was to investigate remote sensing methods to effectively determine the presence of basalt rock outcrops. Five Landsat 5 TM scenes (path 39, row 29) over the year 2007 growing season were processed and analyzed to detect and quantify basalt outcrops across the Clark Area Soil Survey, ID, USA (4,570 km 2 ). The Robust Classification Method (RCM) using the Spectral Angle Mapper (SAM) method and Random Forest (RF) classifications was applied to individual scenes and to a multitemporal stack of the five images. The highest performing RCM basalt classification was obtained using the 18 July scene, which yielded an overall accuracy of 60.45%. The RF classifications applied to the same datasets yielded slightly better overall classification rates when using the multitemporal stack (72.35%) than when using the 18 July scene (71.13%) and the same rate of successfully predicting basalt (61.76%) using out-of-bag sampling. For optimal RCM and RF classifications, uncertainty tended to be lowest in irrigated areas; however, the RCM uncertainty map included more extensive areas of low uncertainty that also encompassed forested hillslopes and riparian areas. RCM uncertainty was sensitive to the influence of bright soil reflectance, while RF uncertainty was sensitive to the influence of shadows. Quantification of basalt requires continued investigation to reduce the influence

Spectroscopic Detection of Nitrogen Concentrations in Sagebrush

The ability to estimate foliar nitrogen in semi-arid landscapes can yield information on nutritional status and improve our limited understanding of controls on canopy photosynthesis. We examined two spectroscopic methods for estimating sagebrush dried leaf and live shrub nitrogen content: first derivative reflectance (FDR) and continuum removal. Both methods used partial least squares (PLS) regression to select wavebands most significantly correlated with nitrogen concentrations in the samples. Sagebrush dried leaf spectra produced PLS models (R 2 = 0.76–0.86) that could predict nitrogen concentrations within the data set more accurately than PLS models generated from live shrub spectra (R 2 = 0.41–0.63). Inclusion of wavelengths associated with leaf water in the FDR transformations appeared to improve regression results. These findings are encouraging and warrant further exploration into sagebrush reflectance spectra to characterize nitrogen concentrations.

Aerodynamic Roughness Length Estimation with Lidar and Imaging Spectroscopy in a Shrub-Dominated Dryland

The aerodynamic roughness length (Z0m) serves an important role in the flux exchange between the land surface and atmosphere. In this study, airborne lidar (ALS), terrestrial lidar (TLS), and imaging spectroscopy data were integrated to develop and test two approaches to estimate Z0m over a shrub dominated dryland study area in south-central Idaho, USA. Sensitivity of the two parameterization methods to estimate Zom was analyzed. The comparison of eddy covariancederived Z0m and remote sensing-derived Z 0m sho1iVed that the accuracy of the estimated Z0m heavily depends on the estimation model and the representation of shrub (e.g., Artemisia tridentata subsp. lryomingensis) height in the models. The geometrical method (RA1994) led to 9 percent (-0.5 cm) and 25% (1.1 cm) errors at site 1 and site 2, respectively, which performed better than the height variability-based method (MR1994) with bias error of 20 percent and 48 percent at site 1 and site 2, respectively. The RA1994 model resulted in a larger range of Zom than the MR1994 method. We also found that the mean, median and 75th percentiles of heights (H75) from ALS provides the best Z 0m estimates in the MR1994 model, while the mean, median, and MAD (Median Absolute Deviation from Median Height), as well as AAD (Mean Absolute Deviation from Mean Height) heights from ALS provides the best Z0m estimates in the RA1994 model. In addition, the fractional cover of shrub and grass, distinguished with ALS and imaging spectroscopy data, provided the opportunity to estimate the frontal area index at the pixel-level to assess the influence of grass and shrub on Z0m estimates in the RA1994 method. Results indicate that grass had little effect on Z 0m in the RA1994 method. The Z0m estimations were tightly coupled with vegetation height and its local variance for the shrubs. Overall, the results demonstrate that the use of height and fractional cover from remote sensing data are promising for estimating Zom• and thus refining land surface models at regional scales in semiarid shrublands. Aihua Li and Nancy F. Glenn are with the Department of Geoscience, Boise State University, 1920 University Drive, Boise ID 83725 (nancyglenn@bosiestate.edu). Jessica J. Mitchell is with the Depa1tment of Geography and Planning, Appalachian State University, Boone NC. Ivlatthew J. Germino is with the US Geological Survey, Forest and Rangeland Ecosystem Science Center, Boise, ID. Joel B. Sankey is with the US Geological Survey, Grand Canyon Monitoring and Research Center, Flagstaff, AZ. Wenguang Zhao and Richard Allen are with Biological Engineering, University of Idaho, Kimberly, ID. PHOTOGRAMMETRk: ENGINEERING & REMOTE SENSING Introduction The roughness of the land surface plays an important role in the flux exchange between the land surface and atmosphere (Sud et aL, 1988; Prueger et al., 2004). Land surface roughness can be characterized by the aerodynamic roughness length (Z0m), which is the height of roughness elements at which the mean wind speed approaches zero given the extrapolation of the logarithmic wind profile (Garratt, 1992; Kaimal and Finnigan, 1994). In dryland ecosystems, such as semiarid shrublands, the spatial distribution of roughness elements and specifically Z0m are key parameters for physical models of aeolian transport and for estimating dust emissions from wind erosion (Prigent et al., 2005; Sankey et al., 2010; Sankey et al., 2013; Nield et al., 2013; Pelletier and Field, 2016) and for land surface models (Dickinson and Henderson-Sellers, 1988; Jasinski and Crago, 1999). Traditionally, Zom is calculated using the Ivlonin-Obukhov similarity theory (MOST) applied to measurements of horizontal V\rind speed profiles (Garratt, 1994; Kustas et al., 1994). Therefore, Z0m can be obtained through observations by an eddy covariance (EC) system which provides meteorological measurements; however, estimating Zom from EC is restricted to a single value in the source area of the EC tower, and thus EC estimates are limited for regional land surface models (Paul-Limoges et al., 2013). To address this issue, studies have used remotely sensed information, such as scatterometer (Prigent et al., 2005) and bi-directional reflectance (Marticorena et al., 2004) data, along with laser altimeter measurements (Menenti and Ritchie, 1994; De Vries et al., 2003, Colin and Faivre, 2010, Weligepolage et al., 2012) for parameterizing Zom over a local or regional scale. Aerodynamic roughness is influenced by the height, geometry, density and pattern of roughness elements which include vegetation and microand macro-topographic features (Garratt, 1992; Lettau, 1969; Raupach, 1992 and 1994; Shaw and Pereira, 1982). Empirical relationships between Zom and measurable characteristics of roughness elements (e.g., vegetation height, normalized difference vegetation index (NDVI), leaf area index (LAI), frontal area index (FAI, A.1)) have been used to parameterize Z0m over a large sale. For example, NDVI and LAI derived from optical remote sensing have been correlated with Zom (Choudhury and Monteith, 1988; Bastiaanssen, 1995; Jia et al., 2003). In some previous studies, Z0m was assumed as a proportion of roughness element height (i.e., Kustas et al., 1989; Garratt, 1992). The three-dimensional (3D) structure of the lands surface and vegetation, as captured by laser altimetry (or light detection and ranging (lidar)) provides a straightforward measure of Photogrammetric Engineering & Remote Sensing Vol. 83, No. 6, June 2017, pp. 415-427. 0099-1112/17/415-427