Global application of an unoccupied aerial vehicle photogrammetry protocol for predicting aboveground biomass in non-forest ecosystems.

Non-forest ecosystems, dominated by shrubs, grasses and herbaceous plants, provide ecosystem services including carbon sequestration and forage for grazing, and are highly sensitive to climatic changes. Yet these ecosystems are poorly represented in remotely sensed biomass products and are undersampled by in situ monitoring. Current global change threats emphasize the need for new tools to capture biomass change in non-forest ecosystems at appropriate scales. Here we developed and deployed a new protocol for photogrammetric height using unoccupied aerial vehicle (UAV) images to test its capability for delivering standardized measurements of biomass across a globally distributed field experiment. We assessed whether canopy height inferred from UAV photogrammetry allows the prediction of aboveground biomass (AGB) across low-stature plant species by conducting 38 photogrammetric surveys over 741 harvested plots to sample 50 species. We found mean canopy height was strongly predictive of AGB across species, with a median adjusted R 2 of 0.87 (ranging from 0.46 to 0.99) and median prediction error from leave-one-out cross-validation of 3.9%. Biomass per-unit-of-height was similar within but different among, plant functional types. We found that photogrammetric reconstructions of canopy height were sensitive to wind speed but not sun elevation during surveys. We demonstrated that our photogrammetric approach produced generalizable measurements across growth forms and environmental settings and yielded accuracies as good as those obtained from in situ approaches. We demonstrate that using a standardized approach for UAV photogrammetry can deliver accurate AGB estimates across a wide range of dynamic and heterogeneous ecosystems. Many academic and land management institutions have the technical capacity to deploy these approaches over extents of 1-10 ha-1. Photogrammetric approaches could provide much-needed information required to calibrate and validate the vegetation models and satellite-derived biomass products that are essential to understand vulnerable and understudied non-forested ecosystems around the globe.

Provoking a Cultural Shift in Data Quality.

Ecological studies require quality data to describe the nature of ecological processes and to advance understanding of ecosystem change. Increasing access to big data has magnified both the burden and the complexity of ensuring quality data. The costs of errors in ecology include low use of data, increased time spent cleaning data, and poor reproducibility that can result in a misunderstanding of ecosystem processes and dynamics, all of which can erode the efficacy of and trust in ecological research. Although conceptual and technological advances have improved ecological data access and management, a cultural shift is needed to embed data quality as a cultural practice. We present a comprehensive data quality framework to evoke this cultural shift. The data quality framework flexibly supports different collaboration models, supports all types of ecological data, and can be used to describe data quality within both short- and long-term ecological studies.

Evaluation of the automated reference toolset as a method to select reference plots for oil and gas reclamation on Colorado Plateau rangelands.

Rangelands are typically characterized by low precipitation and low biomass which makes them susceptible to disturbance and difficult to reclaim. These characteristics become a management issue when considering the widespread and significant impact of oil and gas development on rangelands. Reclamation from this land use involves the complexities of dealing with multiple state and federal agencies, private landowners, and their sometimes conflicting rules. Reference plots (e.g., nearby undisturbed sites) can help with these issues by providing an objective context for reclamation planning. They are selected to provide a comparison that is similar to a reclamation site in most aspects except for the disturbance activity. This allows for the relative condition of the reclamation site to be determined. Because selection of reference plots is normally expert-driven on a site-by-site basis, it can be time consuming and thus ineffective in helping to meet reclamation goals over large landscapes. The Automated Reference Tool (ART) was developed to improve the efficiency and efficacy of reference plot selection. The ART improves reference plot selection through remote sensing and indicators of land potential by selecting reference plots of similar land potential to the reclamation site based on soil texture, topography, and geology. We evaluated the ART in the context of well-pad reclamation to determine if ART-selected plots were appropriate to use as reference when compared to an existing reference plot network. We applied the ART to reclamation sites managed by the Bureau of Land Management's (BLM) White River Field Office, Colorado which had existing expert-selected reference plots. We found that the ART-selected reference plots and their matching expert-selected reference plot had similar large-scale vegetative cover characteristics (total foliar: R2 = 0.34, p-value = 0.0012) and dissimilar finer-scale cover characteristics (plant diversity: R2 = 0.079, p-value = 0.15). In addition, we detected similarities in their soil water content (R2 = 0.43, p-value<0.001), depth to restricting layer (RMSD = 21.90), and rock fragment (RMSD = 19.99). These results demonstrate that ART could be a useful tool for managers to help meet their reclamation goals over large landscapes, but it is not a complete automation of the reference selection process.

Integrating drone imagery with existing rangeland monitoring programs.

The recent availability of small and low-cost sensor carrying unmanned aerial systems (UAS, commonly known as drones) coupled with advances in image processing software (i.e., structure from motion photogrammetry) has made drone-collected imagery a potentially valuable tool for rangeland inventory and monitoring. Drone-imagery methods can observe larger extents to estimate indicators at landscape scales with higher confidence than traditional field sampling. They also have the potential to replace field methods in some instances and enable the development of indicators not measurable from the ground. Much research has already demonstrated that several quantitative rangeland indicators can be estimated from high-resolution imagery. Developing a suite of monitoring methods that are useful for supporting management decisions (e.g., repeatable, cost-effective, and validated against field methods) will require additional exploration to develop best practices for image acquisition and analytical workflows that can efficiently estimate multiple indicators. We embedded with a Bureau of Land Management (BLM) field monitoring crew in Northern California, USA to compare field-measured and imagery-derived indicator values and to evaluate the logistics of using small UAS within the framework of an existing monitoring program. The unified workflow we developed to measure fractional cover, canopy gaps, and vegetation height was specific for the sagebrush steppe, an ecosystem that is common in other BLM managed lands. The correspondence between imagery and field methods yielded encouraging agreement while revealing systematic differences between the methods. Workflow best practices for producing repeatable rangeland indicators is likely to vary by vegetation composition and phenology. An online space dedicated to sharing imagery-based workflows could spur collaboration among researchers and quicken the pace of integrating drone-imagery data within adaptive management of rangelands. Though drone-imagery methods are not likely to replace most field methods in large monitoring programs, they could be a valuable enhancement for pressing local management needs.

A hyper-temporal remote sensing protocol for high-resolution mapping of ecological sites.

Ecological site classification has emerged as a highly effective land management framework, but its utility at a regional scale has been limited due to the spatial ambiguity of ecological site locations in the U.S. or the absence of ecological site maps in other regions of the world. In response to these shortcomings, this study evaluated the use of hyper-temporal remote sensing (i.e., hundreds of images) for high spatial resolution mapping of ecological sites. We posit that hyper-temporal remote sensing can provide novel insights into the spatial variability of ecological sites by quantifying the temporal response of land surface spectral properties. This temporal response provides a spectral 'fingerprint' of the soil-vegetation-climate relationship which is central to the concept of ecological sites. Consequently, the main objective of this study was to predict the spatial distribution of ecological sites in a semi-arid rangeland using a 28-year time series of normalized difference vegetation index from Landsat TM 5 data and modeled using support vector machine classification. Results from this study show that support vector machine classification using hyper-temporal remote sensing imagery was effective in modeling ecological site classes, with a 62% correct classification. These results were compared to Gridded Soil Survey Geographic database and expert delineated maps of ecological sites which had a 51 and 89% correct classification, respectively. An analysis of the effects of ecological state on ecological site misclassifications revealed that sites in degraded states (e.g., shrub-dominated/shrubland and bare/annuals) had a higher rate of misclassification due to their close spectral similarity with other ecological sites. This study identified three important factors that need to be addressed to improve future model predictions: 1) sampling designs need to fully represent the range of both within class (i.e., states) and between class (i.e., ecological sites) spectral variability through time, 2) field sampling protocols that accurately characterize key soil properties (e.g., texture, depth) need to be adopted, and 3) additional environmental covariates (e.g. terrain attributes) need to be evaluated that may help further differentiate sites with similar spectral signals. Finally, the proposed hyper-temporal remote sensing framework may provide a standardized approach to evaluate and test our ecological site concepts through examining differences in vegetation dynamics in response to climatic variability and other drivers of land-use change. Results from this study demonstrate the efficacy of the hyper-temporal remote sensing approach for high resolution mapping of ecological sites, and highlights its utility in terms of reduced cost and time investment relative to traditional manual mapping approaches.

Breaks in MODIS time series portend vegetation change: verification using long-term data in an arid grassland ecosystem.

Frequency and severity of extreme climatic events are forecast to increase in the 21st century. Predicting how managed ecosystems may respond to climatic extremes is intensified by uncertainty associated with knowing when, where, and how long effects of extreme events will be manifest in an ecosystem. In water-limited ecosystems with high inter-annual variability in rainfall, it is important to be able to distinguish responses that result from seasonal fluctuations in rainfall from long-term directional increases or decreases in precipitation. A tool that successfully distinguishes seasonal from directional biomass responses would allow land managers to make informed decisions about prioritizing mitigation strategies, allocating human resource monitoring efforts, and mobilizing resources to withstand extreme climatic events. We leveraged long-term observations (2000-2013) of quadrat-level plant biomass at multiple locations across a semiarid landscape in southern New Mexico to verify the use of Normalized Difference Vegetation Index (NDVI) time series derived from 250-m Moderate Resolution Imaging Spectroradiometer (MODIS) data as a proxy for changes in aboveground productivity. This period encompassed years of sustained drought (2000-2003) and record-breaking high rainfall (2006 and 2008) followed by subsequent drought years (2011 through 2013) that resulted in a restructuring of plant community composition in some locations. Our objective was to decompose vegetation patterns derived from MODIS NDVI over this period into contributions from (1) the long-term trend, (2) seasonal cycle, and (3) unexplained variance using the Breaks for Additive Season and Trend (BFAST) model. BFAST breakpoints in NDVI trend and seasonal components were verified with field-estimated biomass at 15 sites that differed in species richness, vegetation cover, and soil properties. We found that 34 of 45 breaks in NDVI trend reflected large changes in mean biomass and 16 of 19 seasonal breaks accompanied changes in the contribution to biomass by perennial and/or annual grasses. The BFAST method using satellite imagery proved useful for detecting previously reported ground-based changes in vegetation in this arid ecosystem. We demonstrate that time series analysis of NDVI data holds potential for monitoring landscape condition in arid ecosystems at the large spatial scales needed to differentiate responses to a changing climate from responses to seasonal variability in rainfall.

Modeling vegetation heights from high resolution stereo aerial photography: an application for broad-scale rangeland monitoring.

Vertical vegetation structure in rangeland ecosystems can be a valuable indicator for assessing rangeland health and monitoring riparian areas, post-fire recovery, available forage for livestock, and wildlife habitat. Federal land management agencies are directed to monitor and manage rangelands at landscapes scales, but traditional field methods for measuring vegetation heights are often too costly and time consuming to apply at these broad scales. Most emerging remote sensing techniques capable of measuring surface and vegetation height (e.g., LiDAR or synthetic aperture radar) are often too expensive, and require specialized sensors. An alternative remote sensing approach that is potentially more practical for managers is to measure vegetation heights from digital stereo aerial photographs. As aerial photography is already commonly used for rangeland monitoring, acquiring it in stereo enables three-dimensional modeling and estimation of vegetation height. The purpose of this study was to test the feasibility and accuracy of estimating shrub heights from high-resolution (HR, 3-cm ground sampling distance) digital stereo-pair aerial images. Overlapping HR imagery was taken in March 2009 near Lake Mead, Nevada and 5-cm resolution digital surface models (DSMs) were created by photogrammetric methods (aerial triangulation, digital image matching) for twenty-six test plots. We compared the heights of individual shrubs and plot averages derived from the DSMs to field measurements. We found strong positive correlations between field and image measurements for several metrics. Individual shrub heights tended to be underestimated in the imagery, however, accuracy was higher for dense, compact shrubs compared with shrubs with thin branches. Plot averages of shrub height from DSMs were also strongly correlated to field measurements but consistently underestimated. Grasses and forbs were generally too small to be detected with the resolution of the DSMs. Estimates of vertical structure will be more accurate in plots having low herbaceous cover and high amounts of dense shrubs. Through the use of statistically derived correction factors or choosing field methods that better correlate with the imagery, vegetation heights from HR DSMs could be a valuable technique for broad-scale rangeland monitoring needs.

Rangeland and pasture monitoring: an approach to interpretation of high-resolution imagery focused on observer calibration for repeatability.

Collection of standardized assessment and monitoring data is critically important for supporting policy and management at local to continental scales. Remote sensing techniques, including image interpretation, have shown promise for collecting plant community composition and ground cover data efficiently. More work needs to be done, however, evaluating whether these techniques are sufficiently feasible, cost-effective, and repeatable to be applied in large programs. The goal of this study was to design and test an image-interpretation approach for collecting plant community composition and ground cover data appropriate for local and continental-scale assessment and monitoring of grassland, shrubland, savanna, and pasture ecosystems. We developed a geographic information system image-interpretation tool that uses points classified by experts to calibrate observers, including point-by-point training and quantitative quality control limits. To test this approach, field data and high-resolution imagery (∼3 cm ground sampling distance) were collected concurrently at 54 plots located around the USA. Seven observers with little prior experience used the system to classify 300 points in each plot into ten cover types (grass, shrub, soil, etc.). Good agreement among observers was achieved, with little detectable bias and low variability among observers (coefficient of variation in most plots <0.5). There was a predictable relationship between field and image-interpreter data (R (2) > 0.9), suggesting regression-based adjustments can be used to relate image and field data. This approach could extend the utility of expensive-to-collect field data by allowing it to serve as a validation data source for data collected via image interpretation.