In the sporadic permafrost zone of North America, thaw-induced boreal forest loss is leading to permafrost-free wetland expansion. These land cover changes alter landscape-scale surface properties with potentially large, however, still unknown impacts on regional climates. In this study, we combine nested eddy covariance flux tower measurements with satellite remote sensing to characterize the impacts of boreal forest loss on albedo, eco-physiological and aerodynamic surface properties, and turbulent energy fluxes of a lowland boreal forest region in the Northwest Territories, Canada. Planetary boundary layer modelling is used to estimate the potential forest loss impact on regional air temperature and atmospheric moisture. We show that thaw-induced conversion of forests to wetlands increases albedo, bulk surface conductance for water vapour and aerodynamic surface temperature. At the same time, heat transfer efficiency is reduced. These shifts in land surface properties increase latent at the expense of sensible heat fluxes, thus, drastically reducing Bowen ratios. Due to the lower albedo of forests and their masking effect of highly reflective snow, available energy is lower in wetlands, especially in late winter. Modelling results demonstrate that a conversion of a present-day boreal forest–wetland to a hypothetical homogeneous wetland landscape could induce a near-surface cooling effect on regional air temperatures of up to 3–4 °C in late winter and 1–2 °C in summer. An atmospheric wetting effect in summer is indicated by a maximum increase in water vapour mixing ratios of 2 mmol mol−1. At the same time, maximum boundary layer heights are reduced by about a third of the original height. In fall, simulated air temperature and atmospheric moisture between the two scenarios do not differ. Therefore, permafrost thaw-induced boreal forest loss may modify regional precipitation patterns and slow down regional warming trends.
This study demonstrates linkages between the 1997/98 El Niño/Southern Oscillation index and a threshold shift to increased permafrost loss within a southern Taiga Plains watershed, Northwest Territories, Canada. Three-dimensional contraction of permafrost plateaus and changes in vegetation structural characteristics are determined from multi-temporal airborne Light Detection And Ranging (LiDAR) surveys in 2008, 2011 and 2015. Morphological changes in permafrost cover are compared with optical image analogues from 1970, 1977, 2000, and 2008 and time-series hydro-climate data. Results demonstrate that significant changes in air temperature, precipitation, runoff, and a shortening of the snow-covered season by 35 days (1998-2014), and 50 days (1998 only) occurred after 1997. The albedo reduction associated with 35 and 50 days less snow cover leads to increases in shortwave energy receipt during the active thaw period of ~12% (3% annually) and ~16% (5% annually), respectively. From 2000 to 2015, sporadic permafrost loss accelerated from 0.19% (of total basin area) per year between 1970-2000 to 0.58% per year from 2000 to 2015, with a projected total loss of permafrost by ~2044. From ~1997 to 2011, we observe a corresponding shift to increased runoff ratio. However, observed increases in the proportion of snow precipitation and the volumetric contribution of permafrost loss to runoff post-1997 (0.6 to 6.4% per year) cannot fully explain this shift. This suggests increases in drainage efficiency and possible losses from long-term groundwater storage as a result of subtle terrain morphological and soil zone hydraulic conductivity changes. These hydrological changes appear coincident with high vegetation mortality at plateau margins combined with succession-related canopy growth in some bog and fen areas, which are presumed to be drying. Similar changes in runoff response were observed at adjacent Birch, Trout and Jean-Marie River watersheds indicating that observations are representative of northern Boreal sporadic permafrost/wetland watersheds in the Taiga Plains.
Airborne LiDAR is increasingly used in forest carbon, ecosystem, and resource monitoring. For practical design and manufacture reasons, the 1064 nm near-infrared (NIR) wavelength has been the most commonly adopted, and most literature in this field represents sampling characteristics in this wavelength. However, due to eye-safety and application-specific needs, other common wavelengths are 1550 nm and 532 nm. All provide canopy structure reconstructions that can be integrated or compared through space and time but the consistency or complementarity of 3D airborne LiDAR data sampled at multiple wavelengths is poorly understood. Here, we report on multispectral LiDAR missions carried out in 2013 and 2015 over a managed forest research site. The 1st used 3 independent sensors, and the 2nd used a single sensor carrying 3 lasers. The experiment revealed differences in proportions of returns at ground level, vertical foliage distributions, and gap probability across wavelengths. Canopy attenuation was greatest at 532 nm, presumably due to leaf tissue absorption. Relative to 1064 nm, foliage was undersampled at midheight percentiles at 1550 nm and 532 nm. Multisensor data demonstrated differences in foliage characterization due to combined influences of wavelength and acquisition configuration. Single-sensor multispectral data were more stable but demonstrated clear wavelength-dependent variation that could be exploited in intensity-based land cover classification without the aid of 3D derivatives. This work sets the stage for improvements in land surface classification and vertical foliage partitioning through the integration of active spectral and structural laser return information.
The objective of this study is to test a cost-effective, physically based Light Detection and Ranging (LiDAR) classification methodology for wetland and upland land cover types within an area exceeding 1,000 km2 in the Boreal Plains, Alberta, Canada. Decision criteria are based on physical attributes of the landscape that influence maintenance of land cover types. Results are compared with 38 geolocated measurement plots at land cover boundaries and transition zones, manual delineation of 2,337 wetlands using photogrammetric methods and publicly available land cover classifications.
Results suggest that 57% of LiDAR-based wetland classes correspond with delineated wetlands, whereas 37% occur as errors of commission due to excluded wetlands in the manual delineation and confusion with harvested areas. Comparison of classified edges with plot shows that all classifications underestimate wetland area. Residual differences of the LiDAR-based classification are −0.3 m, on average (compared with measured), and have reduced range of error compared with other methods. Multispectral classifications misclassify up to 2/3 of wetland boundaries as a result of lower-resolution mixed pixels. Therefore, high-resolution maps of terrain morphology and vegetation structure provide an accurate, cost-effective means for characterizing wetland vs. upland forest in areas where LiDAR data are available.
At the southern margin of permafrost in North America, climate change causes widespread permafrost thaw. In boreal lowlands, thawing forested permafrost peat plateaus (‘forest’) lead to expansion of permafrost-free wetlands (‘wetland’). Expanding wetland area with saturated and warmer organic soils is expected to increase landscape methane (CH4) emissions. Here, we quantify the thaw-induced increase in CH4 emissions for a boreal forest-wetland landscape in the southern Taiga Plains, Canada, and evaluate its impact on net radiative forcing relative to potential long-term net carbon dioxide (CO2) exchange. Using nested wetland and landscape eddy covariance net CH4 flux measurements in combination with flux footprint modeling, we find that landscape CH4 emissions increase with increasing wetland-to-forest ratio. Landscape CH4emissions are most sensitive to this ratio during peak emission periods, when wetland soils are up to 10 °C warmer than forest soils. The cumulative growing season (May–October) wetland CH4emission of ~13 g CH4 m−2 is the dominating contribution to the landscape CH4 emission of ~7 g CH4 m−2. In contrast, forest contributions to landscape CH4 emissions appear to be negligible. The rapid wetland expansion of 0.26 ± 0.05% yr−1 in this region causes an estimated growing season increase of 0.034 ± 0.007 g CH4 m−2 yr−1 in landscape CH4 emissions. A long-term net CO2 uptake of >200 g CO2 m−2 yr−1 is required to offset the positive radiative forcing of increasing CH4emissions until the end of the 21st century as indicated by an atmospheric CH4 and CO2concentration model. However, long-term apparent carbon accumulation rates in similar boreal forest-wetland landscapes and eddy covariance landscape net CO2 flux measurements suggest a long-term net CO2 uptake between 49 and 157 g CO2 m−2 yr−1. Thus, thaw-induced CH4 emission increases likely exert a positive net radiative greenhouse gas forcing through the 21st century.
This study presents a comparison and integration of three methods commonly used to estimate the amount of forest ecosystem carbon (C) available for storage. In particular, we examine the representation of living above- and below-ground biomass change (net accumulation) using plot-level biometry and repeat airborne laser scanning (ALS) of three dimensional forest plot structure. These are compared with cumulative net CO2fluxes (net ecosystem production, NEP) from eddy covariance (EC) over a six-year period within a jack pine chronosequence of four stands (~ 94, 30, 14 and 3 years since establishment from 2005) located in central Saskatchewan, Canada. Combining the results of the two methods yield valuable observations on the partitioning of C within ecosystems. Subtracting total living biomass C accumulation from NEP results in a residual that represents change in soil and litter C storage. When plotted against time for the stands investigated, the curve produced is analogous to the soil C dynamics described in Covington (1981). Here, ALS biomass accumulation exceeds EC-based NEP measured in young stands, with the residual declining with age as stands regenerate and litter decomposition stabilizes. During the 50–70 year age-period, NEP and live biomass accumulation come into balance, with the soil and litter pools of stands 70–100 years post-disturbance becoming a net store of C. Biomass accumulation was greater in 2008–2011 compared to 2005–2008, with the smallest increase in the 94-year-old “old jack pine” stand and greatest in the 14-year-old “harvested jack pine 1994” stand, with values of 1.4 (± 3.2) tC ha− 1 and 12.0 (± 1.6) tC ha− 1, respectively. The efficiency with which CO2 was stored in accumulated biomass was lowest in the youngest and oldest stands, but peaked during rapid regeneration following harvest (14-year-old stand). The analysis highlights that the primary source of uncertainty in the data integration workflow is in the calculation of biomass expansion factors, and this aspect of the workflow needs to be implemented with caution to avoid large error propagations. We suggest that the adoption of integrated ALS, in situ and atmospheric flux monitoring frameworks is needed to improve spatio-temporal partitioning of C balance components at sub-decadal scale within rapidly changing forest ecosystems and for use in national carbon accounting programs.
Forest soils are generally considered to be net sinks of methane (CH4), but CH4 fluxes vary spatially depending on soil conditions. Measuring CH4 exchange with chambers, which are commonly used for this purpose, might not result in representative fluxes at site scale. Appropriate methods for upscaling CH4 fluxes from point measurements to site scale are therefore needed. At the boreal forest research site, Norunda, chamber measurements of soils and vegetation indicate that the site is a net sink of CH4, while tower gradient measurements indicate that the site is a net source of CH4. We investigated the discrepancy between chamber and tower gradient measurements by upscaling soil CH4 exchange to a 100 ha area based on an empirical model derived from chamber measurements of CH4 exchange and measurements of soil moisture, soil temperature and water table depth. A digital elevation model (DEM) derived from high-resolution airborne Light Detection and Ranging (LiDAR) data was used to generate gridded water table depth and soil moisture data of the study area as input data for the upscaling. Despite the simplistic approach, modeled fluxes were significantly correlated to four out of five chambers with R > 0.68. The upscaling resulted in a net soil sink of CH4of −10 μmol m−2 h−1, averaged over the entire study area and time period (June–September, 2010). Our findings suggest that additional contributions from CH4 soil sources outside the upscaling study area and possibly CH4 emissions from vegetation could explain the net emissions measured by tower gradient measurements.
Quantifying variability of forested riparian buffer (FRB) vegetation structure with variation in adjacent land use supports an understanding of how anthropogenic disturbance influences the ability of riparian systems to perform ecosystem services. However, quantifying FRB structure over large regions is a challenge and requires efficient data collection and processing methods that integrate conventional in situ vegetation sampling with remote sensing data. This study uses automated algorithms to process airborne light detection and ranging (LiDAR) data for mapping of riparian vegetation height, canopy cover and corridor width along 5,900 transects using methods validated in 80 mensuration plots in central Pennsylvania, USA. The key objective of this study was to use airborne LiDAR data to quantify differences in edge vs interior vegetation structure as influenced by buffer width and adjacent land use type, continuously throughout a watershed. Riparian vegetation height, canopy cover and buffer width were estimated along FRB transects adjacent to developed (residential/commercial and agricultural) and undeveloped (grassland) land use types and compared to reference transects within larger forested areas and thus without an edge. On average, buffers adjacent to developed land use types were narrower than those adjacent to natural, undeveloped land use types. Approximately 50% of streams in the watershed had FRB corridors ≤30 m wide. Only 23% of streams had a corridor width ≥200 m, the width recommended to support key ecosystem services. Undeveloped land use types contained taller riparian vegetation and wider corridors, whereas developed land use types contained shorter riparian vegetation and narrow FRB corridors. Edge effects also affected vegetation structure. Vegetation height was 5–8 m shorter at the interface between the FRB and the adjacent land use (the matrix) than in the naturally occurring stream edge or in the corridor interior. Canopy cover was not influenced by adjacent land use type or width. This study demonstrates that airborne LiDAR data can be used to accurately map riparian buffer vegetation width, height and canopy cover to support ecological based management of riparian corridors over wide areas.
Mahoney C, Kljun N, Los SO, Chasmer L, Hacker JM, Hopkinson C, North PRJ, Rosette J, van Gorsel E. Slope estimation from ICESat/GLAS. Remote Sensing. 2014;6(10):10051-10069.
The Western Boreal Plain of North Central Alberta comprises a mosaic of wetlands and aspen (Populus tremuloides) dominated uplands where precipitation (P) is normally exceeded by evapotranspiration (ET). As such these systems are highly susceptible to the climatic variability that may upset the balance between P and ET. Above canopy evapotranspiration (ETC) and understory evapotranspiration (ETB) were examined using the eddy covariance technique situated at 25.5 m (7.5 m above tree crown) and 4.0 m above the ground surface, respectively. During the peak period of the growing seasons (green periods), ETC averaged 3.08 mm d−1 and 3.45 mm d−1 in 2005 and 2006, respectively, while ETB averaged 1.56 mm d−1 and 1.95 mm d−1. Early in the growing season, ETB was equal to or greater than ETC once understory development had occurred. However, upon tree crown growth, ETB was lessened due to a reduction in available energy. ETB ranged from 42 to 56% of ETC over the remainder of the snow-free seasons. Vapour pressure deficit (VPD) and soil moisture (θ) displayed strong controls on both ETC and ETB. ETCresponded to precipitation events as the developed tree crown intercepted and held available water which contributed to peak ETC following precipitation events >10 mm. While both ETC and ETB were shown to respond to VPD, soil moisture in the rooting zone is shown to be the strongest control regardless of atmospheric demand. Further, soil moisture and tension data suggest that rooting zone soil moisture is controlled by the redistribution of soil water by the aspen root system.
This study presents a decision-tree (DT) approach to classifying heterogeneous land cover types within a northern watershed located in the zone of discontinuous permafrost using airborne LiDAR and high resolution spectral datasets. Results are compared with a more typically applied supervised classification. Increasing errors in discharge resulting from an inaccurate classification are quantified using a distributed hydrological model.
The hierarchical classification was accurate between 88% and 97% of the validation sub-area, whereas the parallelepiped classification was accurate between 38% and 74% of the same area (despite overall accuracy of ~ 91%, kappa = 0.91). Topographical derivatives were best able to explain variations in land cover types (82% to 96%), whilst spectral and vegetation structural derivatives were less accurate. When compared with field measurements, the hierarchical classification of plateau edges (adjacent to a fen) was within 2 m of measured, 60% of the time, whilst this occurred only 40% of the time when using a spectral classification. When examining the impacts of land cover classification accuracy on modelled discharge, we find that the length of the Hydrological Response Unit defined by the classification (and subject to varying levels of errors) was linearly related to discharge (m3) such that an increase in permafrost plateau area would increase discharge by 26% of the total. The methodology presented in this paper clarifies previous classification and modelling studies using Landsat and IKONOS data for the same basin. This study greatly improves upon past classifications in the same area, furthers our understanding of the distribution of connected bogs and fens (as conveyors of water to the basin outlet) within the watershed, and current spatial extents of rapidly thawing permafrost plateaus, which are critical for better understanding the impacts of climate change on these northern environments.
This study examines the hydrological recovery of two regenerating boreal trembling aspen (Populus tremuloides Michx.) dominated stands and the sensitivity of that regeneration to drought within the first 5 years of establishment. The results indicate that evapotranspiration fluxes and water-use efficiency rebounded quickly as a result of new vegetation foliage growth and wet conditions found within the first 2 years following the harvest. However, a period of dry years had a significant influence on rates of postharvest growth, carbon dioxide (CO2), and water fluxes at these sites. The northern study area (NSA) and southern study area (SSA) were harvested in the winters of 2007 and 2008, respectively. The first and second years of regeneration at the SSA and NSA, respectively, were marked by an early spring thaw and higher-than-normal precipitation, while air temperatures remained slightly above the 30-year normal. During this period, mean measured height of vegetation tripled at both sites, and cumulative evapotranspiration was approximately 60% of that prior to harvest by the end of the second year of growth. By the third year (2009), the NSA became a sink for atmospheric CO2 during the snow free season (days of the year 128–238) despite low precipitation during the latter half of the summer. Volumetric soil moisture content in 2009 was the highest (on average) of the 5 years examined due to heavy snowfall and a late start to the growing season (where air temperatures consistently exceeded 0 °C), resulting in sustained productivity. However, cumulative annual precipitation also declined to 79% and 57% in 2009 and 2010, respectively, of the 30-year normal for that region, leading to significant (lagged) declines in forest productivity at the NSA in 2010 and 2011. This resulted in the site becoming a source of CO2 to the atmosphere during the 2010 and 2011 growing seasons (annual balance was not measured). Throughout the drought period (2009, 2010, and 2011), mean stand height increased by only 15%, 11%, and 14%, respectively, compared with the mean stand height in 2008. Water-use efficiency also declined in 2010 and 2011, whereas differences in light-use efficiency did not vary significantly because foliage was maintained (i.e., leaves did not abscise as a result of drought). The results of this study indicate that regenerating aspen stands are sensitive to drought and respond relatively quickly to changes in the soil moisture regime. This is important because regional drying as a result of predicted climatic changes combined with increased industrial activity may result in significant decline in productivity within these stands over broad regions.
The Western Boreal Plain of north-central Alberta is prone to water-deficit conditions and is hydrologically sensitive to changes in climate, natural resource extraction and disturbance. Accurate measurement and modelling of the main components of the water balance are important for ecosystem and reclamation management; however, the lack of hydro-meteorological instrumentation found within different land cover types makes quantification of changes to the water balance difficult over large areas. Remote sensing data can provide spatial estimates of land cover distribution and leaf area index (LAI) used as inputs into land surface models. However, land surface models can often suffer from inaccuracies as a result of spatial (coarse pixel) and temporal (discrete acquisition) resolutions, mis-classification and inaccurate representation of LAI using remote sensing data. This study uses high-resolution (1 m × 1 m) Light Detection and Ranging-derived vegetation parameters (land cover type, LAI and 3D vegetation frictional influences on air movement) as inputs into the Penman–Monteith evapotranspiration (ET) model along with measured hydro-meteorological variables. Comparison with eddy covariance (EC) measurements indicated that spatially explicit ET estimates at 1 m resolution (over a 5 km × 5 km study area) provided better estimates compared with bulk average ET estimates per land cover type. ET estimates scaled using spatially variable vegetation inputs only underestimated measured fluxes by 2% and 3% at 22·5 and 3 m EC instrumentation towers, respectively. Bulk averaged ET estimates underestimated measured ET by 5% at the 3 m tower and overestimated EC by 7% at the 22·5 m EC tower. Over coarser scales, the error associated with bulk input parameters can lead to error in overall water balance estimation.
Much of the world's boreal forest occurs on permafrost (perennially cryotic ground). As such, changes in permafrost conditions have implications for forest function and, within the zone of discontinuous permafrost (30–80% permafrost in areal extent), distribution. Here, forested peat plateaus underlain by permafrost are elevated above the surrounding permafrost-free wetlands; as permafrost thaws, ground surface subsidence leads to waterlogging at forest margins. Within the North American subarctic, recent warming has produced rapid, widespread permafrost thaw and corresponding forest loss. Although permafrost thaw-induced forest loss provides a natural analogue to deforestation occurring in more southerly locations, we know little about how fragmentation relates to subsequent permafrost thaw and forest loss or the role of changing conditions at the edges of forested plateaus. We address these knowledge gaps by (i) examining the relationship of forest loss to the degree of fragmentation in a boreal peatland in the Northwest Territories, Canada; and (ii) quantifying associated biotic and abiotic changes occurring across forest-wetland transitions and extending into the forested plateaus (i.e., edge effects). We demonstrate that the rate of forest loss correlates positively with the degree of fragmentation as quantified by perimeter to area ratio of peat plateaus (edge : area). Changes in depth of seasonal thaw, soil moisture, and effective leaf area index (LAIe) penetrated the plateau forests by 3–15 m. Water uptake by trees was sevenfold greater in the plateau interior than at the edges with direct implications for tree radial growth. A negative relationship existed between LAIe and soil moisture, suggesting that changes in vegetation physiological function may contribute to changing edge conditions while simultaneously being affected by these changes. Enhancing our understanding of mechanisms contributing to differential rates of permafrost thaw and associated forest loss is critical for predicting future interactions between the land surface processes and the climate system in high-latitude regions.
Estimates of canopy height (H) and fractional canopy cover (FC) derived from lidar data collected during leaf-on and leaf-off conditions are compared with field measurements from 80 forested riparian buffer plots. The purpose is to determine if existing lidar data flown in leaf-off conditions for applications such as terrain mapping can effectively estimate forested riparian buffer H and FC within a range of riparian vegetation types. Results illustrate that: 1) leaf-off and leaf-on lidar percentile estimates are similar to measured heights in all plots except those dominated by deciduous compound-leaved trees where lidar underestimates H during leaf off periods; 2) canopy height models (CHMs) underestimate H by a larger margin compared to percentile methods and are influenced by vegetation type (conifer needle, deciduous simple leaf or deciduous compound leaf) and canopy height variability, 3) lidar estimates of FC are within 10% of plot measurements during leaf-on periods, but are underestimated during leaf-off periods except in mixed and conifer plots; and 4) depth of laser pulse penetration lower in the canopy is more variable compared to top of the canopy penetration which may influence within canopy vegetation structure estimates. This study demonstrates that leaf-off lidar data can be used to estimate forested riparian buffer canopy height within diverse vegetation conditions and fractional canopy cover within mixed and conifer forests when leaf-on lidar data are not available.
Terrestrial laser scanning (TLS) with the Echidna Validation Instrument (EVI) provides an effective and accurate method for calibrating multiple-return airborne laser scanning (ALS) point cloud distributions to map effective leaf area index (LAIe) and foliage profile within a 1-km diameter test site of mature eucalyptus forest at the Tumbarumba research site, New South Wales, Australia. Plot-based TLS foliage profiles are used as training datasets for the derivation of a scaling function applied to calibrate effective leaf area index (LAIe) from a coincident ALS point cloud. The results of this study show that: a) the mean proportion of the total number of returns within 11.3 m radius of the TLS scan station was 64%. Increasing the radius decreased the level of detail due to occlusion; b) the relationship between TLS LAIe profile and ALS foliage percentile distribution (PD) using all, primary and secondary returns are not linearly related; and c) regressions between TLS LAIe profile and ALS PD, demonstrate better correspondence using a 5th order polynomial applied to all returns (r2 = 0.95; SE = 0.09 m2 m− 2) than aquasi-physically-based Weibull scaling function. The calibration routine was applied to ALS data within a GIS environment to create a 500 m radius 3D map of LAIe. This localised 3D calibration of LAIe was then used as the basis to calculate the overhead canopy extinction coefficient parameter (k), and thereby facilitate upscaling of spatial LAIe estimates to larger domains using a Beer Lambert Law assumption.
As airborne laser scanning (ALS) gains wider adoption to support forest operations in Canada, the consistency and quality of derivative products that support long-term monitoring and planning are becoming a key issues for managers. The Canadian Consortium for Lidar Environmental Applications Research (C-CLEAR) has supported almost 200 projects across Canada since 2000, with forest-related studies being a dominant theme. In 2010 and 2011, field operations were mobilized to support 13 ALS projects spanning almost the full longitudinal gradient of Canada’s forests. This paper presents case studies for seven plus an overview of some best practices and data processing workflow tools that have resulted from these consortium activities. Although the projects and research teams are spread across Canada, the coordination and decade of experience provided through C-CLEAR have brought common methodological elements to all. It is clear that operational, analytical and reporting guidelines that adhere to community accepted standards are required if the benefits promised by ALS forestry are to be realized. A national Lidar Institute that builds upon the C-CLEAR model and focuses on developing standards, guidelines, and certified training would address this need.
Variability of midday net ecosystem CO2 exchange (NEE) and respiration was measured using a transect of closed system chambers spanning transitions from channel fen, permafrost plateau, and ombrotrophic flat bog land cover types during the spring melt season (26 April—6 June 2008). The primary objective was to compare fluxes from different land cover types and topographic variability within zones adjacent to and including rapid permafrost thaw. During this period, the bog was the greatest net source of CO2 to the atmosphere, followed by plateau, and fen. NEE was slightly positive (indicating CO2 loss to the atmosphere) during the snowmelt period (average = 0.009 ± 0.004 mg CO2 m-2 s-1), and increased to 0.025 ± 0.012 mg CO2 m-2 s—1, on average, possibly due to soil thaw and increased microbial activity within two days of completely snow-free conditions. Near surface soil temperature and depth to the water table were the most significant controls of soil and ground cover CO2 fluxes within chambers at all sites (p < 0.05). Analysis of historical aerial photographs and satellite imagery of the area from 1947 to 2008 indicates that plateaus are converting more rapidly into bogs than fen, where 73% of plateau areas (since 1970) that thawed had become bogs (as opposed to 27% conversion into fen). Future research requires establishment of a full ecosystem or land cover greenhouse gas and soil nutrient exchange/transfer program, including CO2and water fluxes as well as dissolved organic and inorganic C, and CH4 losses from the soil. These results contribute to a better understanding of northern soil and ground-cover carbon exchanges as greater areas of permafrost plateaus collapse and form bogs.
This study examines the links between the spatial distribution of three-dimensional vegetation structural characteristics and historical permafrost plateau area changes using airborne light detection and ranging and aerial photography. The results show that vegetation is prone to reduced canopy fractional cover (by up to 50%) and reduced canopy heights (by 16−30%) at the edges of plateaus. Reduced biomass may cause a positive feedback, whereby diminished within- and below-canopy shadowing (by 1 h of shadow time per day) results in increased radiation incident on the ground surface (16% greater at open- vs closed-canopy plateau sites) and increased longwave radiation losses (74% greater at open- vs closed-canopy plateau sites). Increased incident shortwave radiation may result in augmented thawing of permafrost and increased meltwater runoff, which further inhibits vegetation and permafrost persistence. Edge influences on ground thaw cause vegetation to die over several years (confirmed using historical aerial photography), thereby exacerbating thaw and plateau area reduction (plateau area reduction = ~27% over 60 years). Permafrost degradation is also evidenced by the increasingly fragmented characteristics of the landscape.
In this study, a Boolean classification was applied using novel methods to 3-D vegetation structural and topographic attributes found within flux footprint source/sink areas measured by eddy covariance instrumentation. The purpose was to determine if the spatial frequency of 3-D attributes, such as canopy height, effective leaf area index, etc., found within 1 km resolution Moderate Resolution Imaging Spectroradiometer (MODIS) pixels were significantly different from or similar to attributes sampled by flux footprints originating from prevailing wind directions. A Kolmogorov-Smirnov test was used for the first time to apply confidence limits to individual MODIS pixels based on (1) the spatial distribution of cumulative frequencies of attributes representative of those sampled by eddy covariance and (2) temporal representation of MODIS pixels related to area sampling frequency by eddy covariance based on wind direction. Structural and topographic attributes at homogeneous Southern Old Aspen and heterogeneous Upland Aspen sites are representative of 56% and 69% of a 1 km radius area surrounding the tower and 21% and 47% of a 4 × 4 km area. Attributes found within the MODIS “tower” pixel compare well with attributes most frequently sampled by eddy covariance instruments at both sites. By classifying pixels using the Boolean approach, correspondence between MODIS pixels and eddy covariance estimates of gross primary production (GPP) explain up to 13% more variance than using pixels proximal to the tower. This study, therefore, provides a method for choosing MODIS pixels that have similar attributes to those found within footprints most frequently sampled by eddy covariance.
Climate warming and human disturbance in north-western Canada have been accompanied by degradation of permafrost, which introduces considerable uncertainty to the future availability of northern freshwater resources. This study demonstrates the rate and spatial pattern of permafrost loss in a region that typifies the southern boundary of permafrost. Remote-sensing analysis of a 1·0 km2 area indicates that permafrost occupied 0·70 km2 in 1947 and decreased with time to 0·43 km2 by 2008. Ground-based measurements demonstrate the importance of horizontal heat flows in thawing discontinuous permafrost, and show that such thaw produces dramatic land-cover changes that can alter basin runoff production in this region. A major challenge to northern water resources management in the twenty-first century therefore lies in predicting stream flows dynamically in the context of widely occurring permafrost thaw. The need for appropriate water resource planning, mitigation, and adaptation strategies is explained.
The discontinuous permafrost zone has been subject to increased air temperatures over recent decades. Permafrost thaw can cause changes to topography, hydrology, vegetation, and trace gas fluxes, and thus it is important to monitor changes in permafrost area through time. Optical imagery can be used to generate time-series databases of near-surface spectra that may be related to permafrost area. This provides a spatial perspective on area permafrost change that is not easily obtained from field data alone. This study examines the cumulative maximum and minimum errors of aerial and satellite imagery used for change detection within the Scotty Creek watershed, Fort Simpson, NWT, Canada. The results illustrate that, unless unchanging linear features are found throughout every image used (e.g., to be used as multitemporal tie points) and radiometric normalization can be applied (problematic for film images), direct image to image comparisons (e.g., subtraction) are not appropriate. Further, measureable cumulative errors are often produced by misclassification of edges, resolution limitations, and increased landscape fragmentation. At Scotty Creek, increased fragmentation of permafrost plateaus occurred from 1947 to 2008. Cumulative maximum and minimum errors result in an approximate 8%–26% error in permafrost area when compared with the total area of the site. Rates of permafrost area reduction within the study area were approximately 0.5% every year, determined from linear correlation (r2 = 0.91, n = 5). Therefore, based on the maximum cumulative error (a worst-case scenario), approximately 21–32 years (for resolutions of 0.18–1.10 m) is required between images to approximate change within this particular site. Increased (decreased) rates of change at other sites will decrease (increase) the timing required to identify change between images beyond error bounds.
Eddy covariance (EC) measurements have greatly advanced our knowledge of carbon exchange in terrestrial ecosystems. However, appropriate techniques are required to upscale these spatially discrete findings globally. Satellite remote sensing provides unique opportunities in this respect, but remote sensing of the photosynthetic light-use efficiency (ε), one of the key components of Gross Primary Production, is challenging. Some progress has been made in recent years using the photochemical reflectance index, a narrow waveband index centered at 531 and 570 nm. The high sensitivity of this index to various extraneous effects such as canopy structure, and the view observer geometry has so far prevented its use at landscape and global scales. One critical aspect of upscaling PRI is the development of generic algorithms to account for structural differences in vegetation. Building on previous work, this study compares the differences in the PRI: ɛ relationship between a coastal Douglas-fir forest located on Vancouver Island, British Columbia, and a mature Aspen stand located in central Saskatchewan, Canada. Using continuous, tower-based observations acquired from an automated multi-angular spectro-radiometer (AMSPEC II) installed at each site, we demonstrate that PRI can be used to measure ɛ throughout the vegetation season at the DF-49 stand (r2 = 0.91, p < 0.00) as well as the deciduous site (r2 = 0.88, p < 0.00). It is further shown that this PRI signal can be also observed from space at both sites using daily observations from the Moderate Resolution Imaging Spectro-radiometer (MODIS) and a multi-angular implementation of atmospheric correction (MAIAC) (r2 = 0.54 DF-49; r2 = 0.63 SOA; p < 0.00). By implementing a simple hillshade model derived from airborne light detection and ranging (LiDAR) to approximate canopy shadow fractions (αs), it is further demonstrated that the differences observed in the relationship between PRI and ε at DF-49 and SOA can be attributed largely to differences in αs. The findings of this study suggest that algorithms used to separate physiological from extraneous effects in PRI reflectance may be more broadly applicable and portable across these two climatically and structurally different biome types, when the differences in canopy structure are known.
The influence of digital elevation model (DEM) resolution to modelled glacier melt during peak melt production was evaluated by performing a clear sky GIS radiation simulation over the Peyto Glacier in the Canadian Rockies. DEMs were generated at eight resolutions ranging from 1 m to 1000 m grid spacing from airborne lidar data. When applied to the planar area (PA) of the terrain, it was found that total melt increased with DEM resolution (r2 = 0·63) by 4% over 3 orders of magnitude. This systematic scaling-effect was mitigated at the basin scale, however, when the DEM slope variant area (SVA) was used to account for the increased divergence from PA as resolution increases. However, even after the inclusion of SVA in glacier surface melt simulations, localized melt variations with scale were still evident in the ablation and accumulation zone observations. In the ablation zone, there was a systematic increase in simulated melt (∼4%) as resolution decreased from 1 m to 1000 m (r2 = 0·89), with the opposite effect in the accumulation zone (r2 = 0·81). DEM resolution also affected the diurnal melt cycle, such that for the entire glacier there was a tendency for a morning over-estimation and afternoon underestimation of melt rate with decreasing resolution. For the accumulation zone, there was an increased melt rate at low resolutions occurring in the afternoon, while in the ablation zone there was a tendency for increasing melt rates with decreasing resolution throughout the day. These localized spatio-temporal variations in simulated melt are largely due to the lowering of ridges and raising of valley floors that occur as resolution decreases. This scale dependence in the representation of terrain morphology directly controls the pattern and relative proportion of direct beam shadowing over actively melting surfaces and thereby has a systematic influence on the grid cell-level hydrological balance. It is recommended that GIS-based glacier melt modelling routines take into account the slope area of grid cells, while noting that the choice of DEM scale can have a discernible and systematic influence on modelled runoff magnitude. It is important to note that while higher grid resolutions mitigate the effect of terrain smoothing on spatio-temporal melt patterns, lower resolutions actually mitigate the systematic error associated with assuming all surface areas are planar.
Field studies were initiated in 1999 at Scotty Creek in the lower Liard River basin, NWT, Canada, to improve understanding of and ability to predict the major water fluxes and storage processes within a wetland-dominated zone of the discontinuous permafrost region. This paper synthesises a decade of published and unpublished research at Scotty Creek for the purpose of presenting the major factors that should be considered by water scientists and managers as a basis for modelling and management strategies. Five main topics are covered: (1) peatlands of lower Liard River valley; (2) hydrological characteristics of permafrost plateaus, flat bogs, and channel fens; (3) runoff generation on permafrost plateaus; (4) conceptual model of peatland hydrology; and (5) climate warming and implications for basin runoff. This synthesis offers a practical understanding of the hydrology of wetland-dominated basins with discontinuous permafrost. It also offers insight into how landscape changes resulting from climate or human disturbances may influence the basin hydrograph.
Understanding the influence of within-pixel land cover heterogeneity is essential for the extrapolation of measured and modeled CO2 fluxes from the canopy to regional scales using remote sensing. Airborne light detection and ranging (lidar) was used to estimate spatial and temporal variations of gross primary production (GPP) across a jack pine chronosequence of four sites in Saskatchewan, Canada for comparison with the Moderate Resolution Imaging Spectroradiometer (MODIS) GPP product. This study utilizes high resolution canopy structural information obtained from airborne lidar to bridge gaps in spatial representation between plot, eddy covariance (EC), and MODIS estimates of vegetation GPP. First we investigate linkages between canopy structure obtained from measurements and light response curves at a jack pine chronosequence during the growing season of 2004. Second, we use the measured canopy height and foliage cover inputs to create a structure-based GPP model (GPPLandsberg) which was tested in 2005. The GPP model is then run using lidar data (GPPLidar) and compared with eight-day cumulative MODIS GPP (GPPMODIS) and EC observations (GPPEC). Finally, we apply the lidar GPP model at spatial resolutions of 1 m to 1000 m to examine the influence of within-pixel heterogeneity and scaling (or pixel aggregation) on GPPLidar. When compared over eight-day cumulative periods throughout the 2005 growing season, the standard deviation of differences between GPPlidar and GPPMODIS were less than differences between either of them and GPPEC at all sites. As might be expected, the differences between pixel aggregated GPP estimates are most pronounced at sites with the highest levels of spatial canopy heterogeneity. The results of this study demonstrate one method for using lidar to scale between eddy covariance flux towers and coarse resolution remote sensing pixels using a structure-based Landsberg light curve model.
Variability in three Pacific teleconnection patterns are examined to see if net carbon exchange at a low-elevation, old-growth forest is affected by climatic changes associated with these periodicities. Examined are the Pacific Decadal Oscillation (PDO), Pacific/North American Oscillation (PNA) and El Niño-Southern Oscillation (ENSO). We use 9 years of eddy covariance CO2, H2O and energy fluxes measured at the Wind River AmeriFlux site, Washington, USA and 8 years of tower-pixel remote sensing data from the Moderate Resolution Imaging Spectroradiometer (MODIS) to address this question. We compute a new Composite Climate Index (CCI) based on the three Pacific Oscillations to divide the measurement period into positive- (2003 and 2005), negative- (1999 and 2000) and neutral-phase climate years (2001, 2002, 2004, 2006 and 2007). The forest transitioned from an annual net carbon sink (NEP=+217 g C m−2 yr−1, 1999) to a source (NEP=−100 g C m−2 yr−1, 2003) during two dominant teleconnection patterns. Net ecosystem productivity (NEP), water use efficiency (WUE) and light use efficiency (LUE) were significantly different (P<0.01) during positive (NEP=−0.27 g C m−2 day−1, WUE=4.1 mg C g−1 H2O, LUE=0.94 g C MJ−1) and negative (NEP=+0.37 g C m−2 day−1, WUE=3.4 mg C g−1 H2O, LUE=0.83 g C MJ−1) climate phases. The CCI was linked to variability in the MODIS Enhanced Vegetation Index (EVI) but not to MODIS Fraction of absorbed Photosynthetically Active Radiation (FPAR). EVI was highest during negative climate phases (1999 and 2000) and was positively correlated with NEP and showed potential for using MODIS to estimate teleconnection-driven anomalies in ecosystem CO2 exchange in old-growth forests. This work suggests that any increase in the strength or frequency of ENSO coinciding with in-phase, low frequency Pacific oscillations (PDO and PNA) will likely increase CO2 uptake variability in Pacific Northwest conifer forests.
Four LiDAR-based models of canopy fractional cover (FCLiDAR) have been tested against hemispherical photography fractional cover measurements (FCHP) and compared across five ecozones, eight forest species and multiple LiDAR survey configurations. The four models compared are based on: i) a canopy-to-total first returns ratio (FCLiDAR(FR)) method; ii) a canopy-to-total returns ratio (FCLiDAR(RR)); iii) an intensity return ratio (FCLiDAR(IR)); and iv) a Beer's Law modified (two-way transmission loss) intensity return ratio (FCLiDAR(BL)). It is found that for the entire dataset, the FCLiDAR(RR)model demonstrates the lowest overall predictive capability of overhead FC (annulus rings 1–4) (r2 = 0.70), with a slight improvement for the FCLiDAR(FR) model (r2 = 0.74). The intensity-based FCLiDAR(IR) model displays the best results (r2 = 0.78). However, the FCLiDAR(BL) model is considered generally more useful (r2 = 0.75) because the associated line of best fit passes through the origin, has a slope near unity and produces a mean estimate of FCHP within 5%. Therefore, FCLiDAR(BL) requires the least calibration across a broad range of forest cover types. The FCLiDAR(FR) and FCLiDAR(RR) models, on the other hand, were found to be sensitive to variations in both canopy height and sensor pulse repetition frequency (or pulse power); i.e. changing the repetition frequency led to a systematic shift of up to 11% in the mean FCLiDAR(RR) estimates while it had no effect on the intensity-based FCLiDAR(IR) or FCLiDAR(BL) models. While the intensity-based models were generally more robust, all four models displayed at least some sensitivity to variations in canopy structural class, suggesting that some calibration of FCLiDAR might be necessary regardless of the model used. Short (< 2 m tall) or open canopy forest plots posed the greatest challenge to accurate FC estimation regardless of the model used.
Carbon dioxide, water vapour, and energy fluxes vary spatially and temporally within forested environments. However, it is not clear to what extent they vary as a result of variability in the spatial distribution of biomass and elevation. The following study presents a new methodology for extracting changes in the structural characteristics of vegetation and elevation within footprint areas, for direct comparison with eddy covariance (EC) CO2 flux concentrations. The purpose was to determine whether within-site canopy structure and local elevation influenced CO2 fluxes in a mature jack pine (Pinus banksiana Lamb.) forest located in Saskatchewan, Canada. Airborne light detection and ranging (lidar) was used to extract tree height, canopy depth, foliage cover, and elevation within 30 min flux footprints. Within-footprint mean structural components and elevation were related to 30 min mean net ecosystem productivity (NEP) and gross ecosystem production (GEP). NEP and GEP were modeled using multiple regression, and when compared with measured fluxes, almost all periods showed improvements in the prediction of flux concentration when canopy structure and elevation were included. Increased biomass was related to increased NEP and GEP in June and August when the ecosystem was not limited by soil moisture. On a daily basis, fractional cover and elevation had varying but significant influences on CO2 fluxes.
Light-use efficiency (LUE) is the ability of vegetated canopies to use light for photosynthesis. Together with remote sensing estimates of canopy cover and meteorological inputs, LUE provides a physical basis for scaling carbon uptake processes from the stand to the global scale. A better understanding of the factors that control LUE will result in improved global estimates of carbon uptake from the terrestrial biosphere. To examine factors that control variability in LUE in stands of different ages during dry and wet conditions, we measured LUE in a chronosequence of four jack pine stands (recent clearcut (age 1–3), regenerating (age 8–9), immature (age 29–30) and mature (∼90 years old)) during one normal (2002), one very dry (2003) and two very wet (2004, 2005) growing seasons in Saskatchewan, Canada. Cumulative CO2 fluxes decreased significantly at all sites during the drought year of 2003, as did mean LUE. Canopy foliage at the recently regenerating jack pine site increased by 19% between 2002 and 2003. Foliage growth rate was reduced by 6% between 2003 and 2004, and foliage biomass decreased by 6% from 2004 to 2005. Over the four years studied, LUE was greatest at the mature jack pine site and lower, but similar, at the other three sites. Mean growing-season LUE varied with mean soil water content at each site, except at that of the newly regenerating stand where soil water had little influence. Mean daily vapor pressure deficit typically had the greatest influence on variability in LUE at all sites. Diffuse versus direct radiation also had significant but varying effects on LUE in jack pine stands of different ages.
The purpose of this study was to estimate the fraction of photosynthetically active radiation absorbed by the canopy (fPAR) from point measurements to airborne lidar for hierarchical scaling up and assessment of the Moderate Resolution Imaging Spectroradiometer (MODIS) fPAR product within a “medium-sized” (7 km × 18 km) watershed. Nine sites across Canada, containing one or more (of 11) distinct species types and age classes at varying stages of regeneration and seasonal phenology were examined using a combination of discrete pulse airborne scanning Light Detection And Ranging (lidar) and coincident analog and digital hemispherical photography (HP). Estimates of fPAR were first compared using three methods: PAR radiation sensors, HP, and airborne lidar. HP provided reasonable estimates of fPAR when compared with radiation sensors. A simplified fractional canopy cover ratio from lidar based on the number of within canopy returns to the total number of returns was then compared with fPAR estimated from HP at 486 geographically registered measurement locations. The return ratio fractional cover method from lidar compared well with HP-derived fPAR (coefficient of determination = 0.72, RMSE = 0.11), despite varying the lidar survey configurations, canopy structural characteristics, seasonal phenologies, and possible slight inaccuracies in location using handheld GPS at some sites. Lidar-derived fractional cover estimates of fPAR were ∼ 10% larger than those obtained using HP (after removing wood components), indicating that lidar likely provides a more realistic estimate of fPAR than HP when compared with radiation sensors. Finally, fPAR derived from lidar fractional cover was modelled at 1 m resolution and averaged over 99 1 km areas for comparison with MODIS fPAR. The following study is one of the first to scale between plot measurements and MODIS pixels using airborne lidar.