Neal J. Pastick

Spatial variability and landscape controls of near‐surface permafrost within the Alaskan Yukon River Basin

The distribution of permafrost is important to understand because of permafrosts influence on high-latitude ecosystem structure and functions. Moreover, near-surface (defined here as within 1 m of the Earths surface) permafrost is particularly susceptible to a warming climate and is generally poorly mapped at regional scales. Subsequently, our objectives were to (1) develop the first-known binary and probabilistic maps of near-surface permafrost distributions at a 30 m resolution in the Alaskan Yukon River Basin by employing decision tree models, field measurements, and remotely sensed and mapped biophysical data; (2) evaluate the relative contribution of 39 biophysical variables used in the models; and (3) assess the landscape-scale factors controlling spatial variations in permafrost extent. Areas estimated to be present and absent of near-surface permafrost occupy approximately 46% and 45% of the Alaskan Yukon River Basin, respectively; masked areas (e.g., water and developed) account for the remaining 9% of the landscape. Strong predictors of near-surface permafrost include climatic indices, land cover, topography, and Landsat 7 Enhanced Thematic Mapper Plus spectral information. Our quantitative modeling approach enabled us to generate regional near-surface permafrost maps and provide essential information for resource managers and modelers to better understand near-surface permafrost distribution and how it relates to environmental factors and conditions.

Evidence for nonuniform permafrost degradation after fire in boreal landscapes

Fire can be a significant driver of permafrost change in boreal landscapes, altering the availability of soil carbon and nutrients that have important implications for future climate and ecological succession. However, not all landscapes are equally susceptible to fire-induced change. As fire frequency is expected to increase in the high latitudes, methods to understand the vulnerability and resilience of different landscapes to permafrost degradation are needed. We present a combination of multiscale remote sensing, geophysical, and field observations that reveal details of both near-surface ( 1 m) impacts of fire on permafrost. Along 11 transects that span burned-unburned boundaries in different landscape settings within interior Alaska, subsurface electrical resistivity and nuclear magnetic resonance data indicate locations where permafrost appears to be resilient to disturbance from fire, areas where warm permafrost conditions exist that may be most vulnerable to future change, and also areas where permafrost has thawed. High-resolution geophysical data corroborate remote sensing interpretations of near-surface permafrost and also add new high-fidelity details of spatial heterogeneity that extend from the shallow subsurface to depths of about 10 m. Results show that postfire impacts on permafrost can be variable and depend on multiple factors such as fire severity, soil texture, soil moisture, and time since fire.

Spatially explicit estimation of aboveground boreal forest biomass in the Yukon River Basin, Alaska

Quantification of aboveground biomass (AGB) in Alaska’s boreal forest is essential to the accurate evaluation of terrestrial carbon stocks and dynamics in northern high-latitude ecosystems. Our goal was to map AGB at 30 m resolution for the boreal forest in the Yukon River Basin of Alaska using Landsat data and ground measurements. We acquired Landsat images to generate a 3-year (2008–2010) composite of top-of-atmosphere reflectance for six bands as well as the brightness temperature (BT). We constructed a multiple regression model using field-observed AGB and Landsat-derived reflectance, BT, and vegetation indices. A basin-wide boreal forest AGB map at 30 m resolution was generated by applying the regression model to the Landsat composite. The fivefold cross-validation with field measurements had a mean absolute error (MAE) of 25.7 Mg ha−1 (relative MAE 47.5%) and a mean bias error (MBE) of 4.3 Mg ha−1 (relative MBE 7.9%). The boreal forest AGB product was compared with lidar-based vegetation height data; the comparison indicated that there was a significant correlation between the two data sets.

Spatiotemporal Analysis of Landsat-8 and Sentinel-2 Data to Support Monitoring of Dryland Ecosystems

Drylands are the habitat and source of livelihood for about two fifths of the world’s population and are highly susceptible to climate and anthropogenic change. To understand the vulnerability of drylands to changing environmental conditions, land managers need to effectively monitor rates of past change and remote sensing offers a cost-effective means to assess and manage these vast landscapes. Here, we present a novel approach to accurately monitor land-surface phenology in drylands of the Western United States using a regression tree modeling framework that combined information collected by the Operational Land Imager (OLI) onboard Landsat 8 and the Multispectral Instrument (MSI) onboard Sentinel-2. This highly-automatable approach allowed us to precisely characterize seasonal variations in spectral vegetation indices with substantial agreement between observed and predicted values (R2 = 0.98; Mean Absolute Error = 0.01). Derived phenology curves agreed with independent eMODIS phenological signatures of major land cover types (average r-value = 0.86), cheatgrass cover (average r-value = 0.96), and growing season proxies for vegetation productivity (R2 = 0.88), although a systematic bias towards earlier maturity and senescence indicates enhanced monitoring capabilities associated with the use of harmonized Landsat-8 Sentinel-2 data. Overall, our results demonstrate that observations made by the MSI and OLI can be used in conjunction to accurately characterize land-surface phenology and exclusion of imagery from either sensor drastically reduces our ability to monitor dryland environments. Given the declines in MODIS performance and forthcoming decommission with no equivalent replacement planned, data fusion approaches that integrate observations from multispectral sensors will be needed to effectively monitor dryland ecosystems. While the synthetic image stacks are expected to be locally useful, the technical approach can serve a wide variety of applications such as invasive species and drought monitoring, habitat mapping, production of phenology metrics, and land-cover change modeling.

Geospatial data mining for digital raster mapping

We performed an in-depth literature survey to identify the most popular data mining approaches that have been applied for raster mapping of ecological parameters through the use of Geographic Information Systems (GIS) and remotely sensed data. Popular data mining approaches included decision trees or “data mining” trees which consist of regression and classification trees, random forests, neural networks, and support vector machines. The advantages of each data mining approach as well as approaches to avoid overfitting are subsequently discussed. We also provide suggestions and examples for the mapping of problematic variables or classes, future or historical projections, and avoidance of model bias. Finally, we address the separate issues of parallel processing, error mapping, and incorporation of “no data” values into modeling processes. Given the improved availability of digital spatial products and remote sensing products, data mining approaches combined with parallel processing potentials should greatly improve the quality and extent of ecological datasets.

NMR for Near-surface Investigations (Development and Applications)

In porous materials, susceptibility contrasts between the matrix and the pore fluid generate pore-scale inhomogeneities in the magnetic field that are referred to as internal gradients. Internal gradients impact NMR measurements, and can cause large errors in the calculated diffusion coefficient if they are not accounted for. The magnitude of the internal gradients is determined by the susceptibility contrast, the strength of the background magnetic field, and the pore geometry. We use statistical analysis to look for correlation between measured internal gradients and properties of sediment samples. The primary goal of this analysis was to identify parameters that could be used as predictors of internal gradient magnitudes. We measured internal gradients using two different NMR methods: Method 1 estimates an average effective gradient, and Method 2 calculates a distribution of effective gradients. The sediment properties that we consider are magnetic susceptibility, iron content, specific surface area, grain size, and measured NMR parameters including the mean log T2 and the T1/T2 ratio. In our preliminary analysis, conducted with data from 20 sediment samples, we observe linear trends between iron content and measured gradients, and between magnetic susceptibility and measured gradients. We also see that the mineral form of iron appears to impact the relationships between iron content, magnetic susceptibility, and internal gradients. The correlation observed between gradients measured with Method 1 and both the specific surface area and T2 could indicate that this method is biased by relaxation time; this relationship was not observed for the gradients measured with Method 2. We plan to collect data on more sediment samples to better understand these relationships and develop a model for the estimation of internal gradients. Such a model will enable us to include internal gradient values in diffusion coefficient calculations for a range of nearsurface applications.