Mapping Aboveground and Foliage Biomass Over the Porcupine Caribou Habitat in Northern Yukon and Alaska Using Landsat and JERS-1/SAR Data

The linkage between caribou and the aboriginal people in the North America has existed for thousands of years. Caribou have played a critical role in the economy, culture, and way of life of the aboriginal people (Hall, 1989; Madsen, 2001). Currently, there are 60 major migratary tundra caribou herds circum-arctic, of which 30 are located in North America, including the Porcupine caribou herd in northern Yukon, Canada and northern Alaska, USA (Russell et al., 1992; Russell et al., 1993; Russell & McNeil, 2002; Russell et al., 2002; Griffith et al., 2002). The Porcupine caribou herd has been at the centre of debate between wildlife habitat conservation and industrial development in the Arctic because of the potential oil drilling in the Arctic National Wildlife Refuge (ANWR) 1002 area, which happens to largely overlap with the calving ground of the Porcupine caribou herd (Griffith et al., 2002; Kaiser, 2002; National Research Council, 2003; Heuer, 2006). One of the objectives of the Canadian International Polar Year (IPY) project entitled “Climate Change Impacts on Canadian Arctic Tundra Ecosystems (CiCAT): Interdisciplinary and Multi-scale Assessments” was to assess the impact of climate change on caribou habitats over Canada’s north, in close collaboration with the CircumArctic Ranfiger Monitoring and Assessment network (CARMA) (http://www.rangifer.net /carma/). Because of the vastness and remoteness of the arctic landmass, inherent logistic difficulty, and high cost of conducting field measurements, an approach that is solely based on field inventory is clearly impractical for monitoring and assessing the impact of climate on caribou habitats. Satellite remote sensing can monitor land surfaces from space repeatedly and consistently over large areas. Therefore, remote sensing provides a powerful tool for monitoring and assessing the impact of climate on caribou habitats, when calibrated and validated against the field measurements and other independent data. In this study, we report the development of baseline maps of aboveground and foliage biomass over the Porcupine caribou habitat in northern Yukon and Alaska, using Landsat and JERS1/SAR data. Specifically, we will (1) describe aboveground and foliage biomass measurement, (2) establish and validate relationships between measurements and remote sensing indices, and (3) map aboveground and foliage biomass for the Porcupine caribou habitat.


Introduction
The linkage between caribou and the aboriginal people in the North America has existed for thousands of years.Caribou have played a critical role in the economy, culture, and way of life of the aboriginal people (Hall, 1989;Madsen, 2001).Currently, there are 60 major migratary tundra caribou herds circum-arctic, of which 30 are located in North America, including the Porcupine caribou herd in northern Yukon, Canada and northern Alaska, USA (Russell et al., 1992;Russell et al., 1993;Russell & McNeil, 2002;Russell et al., 2002;Griffith et al., 2002).The Porcupine caribou herd has been at the centre of debate between wildlife habitat conservation and industrial development in the Arctic because of the potential oil drilling in the Arctic National Wildlife Refuge (ANWR) 1002 area, which happens to largely overlap with the calving ground of the Porcupine caribou herd (Griffith et al., 2002;Kaiser, 2002;National Research Council, 2003;Heuer, 2006).One of the objectives of the Canadian International Polar Year (IPY) project entitled "Climate Change Impacts on Canadian Arctic Tundra Ecosystems (CiCAT): Interdisciplinary and Multi-scale Assessments" was to assess the impact of climate change on caribou habitats over Canada's north, in close collaboration with the CircumArctic Ranfiger Monitoring and Assessment network (CARMA) (http://www.rangifer.net/carma/).Because of the vastness and remoteness of the arctic landmass, inherent logistic difficulty, and high cost of conducting field measurements, an approach that is solely based on field inventory is clearly impractical for monitoring and assessing the impact of climate on caribou habitats.Satellite remote sensing can monitor land surfaces from space repeatedly and consistently over large areas.Therefore, remote sensing provides a powerful tool for monitoring and assessing the impact of climate on caribou habitats, when calibrated and validated against the field measurements and other independent data.In this study, we report the development of baseline maps of aboveground and foliage biomass over the Porcupine caribou habitat in northern Yukon and Alaska, using Landsat and JERS-1/SAR data.Specifically, we will (1) describe aboveground and foliage biomass measurement, (2) establish and validate relationships between measurements and remote sensing indices, and (3) map aboveground and foliage biomass for the Porcupine caribou habitat.

www.intechopen.com
Remote Sensing of Biomass -Principles and Applications 232 2. Data sources and methods 2.1 Field measurements of aboveground and foliage biomass Field measurements were made in Yukon and North Western Territories (Fig. 1).Aboveground biomass was measured at 43 sites in the summer of 2004 along the Dempster Highway, which goes through the winter and summer ranges of the Porcupine caribou habitat (Fig. 1).Foliage biomass was measured at 10 non-treed sites along the Dempster Highway in the summer of 2006, and again in the summer of 2008 at 11 non-treed sites in the Ivvavik National Park.The Ivvavik National Park is located at northern tip of the Yukon and overlaps with the calving ground and summer range of the Porcupine caribou herd.Details of measurement procedure, calculation method, and results are described as follows.We selected sites that were representative of local vegetation conditions, relatively homogenous, and of size at least 3 × 3 Landsat pixels (i.e., > 90 m × 90 m), to ensure that the field measurements can be reliably correlated with Landsat-scale remote sensing indices.In the north where the growing season is very short, biophysical parameters (e.g., foliage biomass) vary significantly both during the beginning and end of the growing season, but are relatively stable and achieve their maximums during the middle of the growing season.Therefore, we measured aboveground biomass and foliage biomass during middle of growing season (e.g., July 18-27, 2004, July 19-27, 2006, and July 15-26, 2008).
We used a systematic approach to sample and measure the aboveground biomass: layer by layer from top to the ground.If a site was sparsely treed woodland, we first measured tree biomass with a variable sampling plot scheme using a prism (Halliwell & Apps, 1997).The basic principle of the variable sampling plot scheme is that each tree is selected using a circular plot with a radius r that is proportional to its diameter at the breast height (DBH).
The constant of proportionality is called the plot radius factor (PRF), and thus for tree i we can write: with DBH is in centimeters and plot radius is in meters, PRF is in m cm -1 .A tree is included or "in" if DBH i > r i /PRF, and excluded or "out" otherwise.A difficulty may arise as whether to include or exclude a tree if it's at the edge, namely, DBH i ≈ r i /PRF.If this is the case, the distance from the plot centre to the tree is measured.Using the tree's DBH and the appropriate PRF, this distance can then be compared to the radius of the plot as calculated by equation 1, and the tree place "in" or "out" of the plot after-the-fact.The measurements included tree height, diameter at breast height (DBH), and stand density.The heights of trees are recorded using a laser height measurement instrument (Impulse Forest Pro, Lasertech, Clarkston, MI, USA).Trees for height measurement are selected on basis of the point plot scheme.The value of DBH was measured using a tape.The stocking and biomass of the trees were then calculated on basis of allometric equations.
If tree regeneration was present, we measured the biomass of the seedlings using a fixed circular plot of radius of 3.99 m.The measurements included counting the number of stems per tree species, selecting an average-sized seedling to estimate its ground stem diameter, height, and sampling one or two average-sized seedling to measure its fresh and oven-dry weight.
For the understory layers, namely, high shrub, low shrub, herbs/graminoids, moss, and lichen, we harvested samples at five 1 m × 1 m plots to measure the total aboveground biomass: four at four directions and one random (Fig. 2).All the aboveground biomass within the plots were cut and collected into plastic bags and their green biomass weighed.
Out of the five plots, we selected one representative plot to conduct an intensive sampling.
For this plot, we measured the average height, visually estimate the ground cover percentage, and then cut and weighed the fresh biomass, sequentially, for tall shrub, low shrubs, herbs/graminoids, moss, and lichen.The ratio of green to oven-dry biomass was measured by bringing a sample of the aboveground biomass from the plot back to laboratory.
If a site had no trees, then we applied the aforementioned understory procedure.Note that the plot number and design of a specific site were determined on basis of site conditions: the more heterogeneous and sparse the vegetation is at a site, the more plots are needed (Chen et al., 2009b;Chen et al., 2010b).The procedure was modified for tall shrubs if the tall shrubs (higher than 50 cm) were clustered instead of homogeneous distribution.Our field experience indicated that it was often the case for tall shrubs.In this case, we measured the biomass of tall shrubs using a fixed circular plot of radius = 3.99 m.The measurements included identifying shrub species, counting number of clusters of shrub in the plot, and selecting two average-sized clusters of shrub for more extensive sampling.All aboveground biomass of the two clusters of shrub were cut and weighed for total green biomass.A fraction of the biomass was brought back to measure the ratio of oven-dry to fresh weight.If there are more than one tall shrub species, we repeated the measurements for each of them.
The procedure for low shrub, herbs/graminoids, moss, and lichen remain the same.
For foliage biomass measurement during the summers of 2006 and 2008, we selected only non-treed sites.At each sites, five 1 m × 1 m plots were sampled (Fig. 3).Fig. 3 shows a photograph of the field sampling at a coastal plain site in the Ivvavik National Park during the summer of 2008.At each plot, all plants were harvested, sorted into dead and live, different species, and leaves and stem, and recorded for their fresh weights.A fraction of the harvested biomass for all components were brought back to the laboratory, oven-dried, and weighed.Oven-dry foliage biomass of vascular plants were than calculated using the ratio of oven-dry to fresh foliage biomass and the corresponding fresh weight records.The tree oven-dry aboveground biomass at a woodland site was calculated using Canadawide tree biomass equations by Evert (1985) and the Weibull tree-size distribution function within a stand by Chen (2004).

GPS
In order to determine the tree-size distribution for a woodland site, the Parameter Prediction Method (PPM) based on basal area was implemented (Chen, 2004).With measured average DBH, height, and stand density (trees ha -1 ) as inputs, the tree-size distribution function gives 15 tree-size classes for each site, and output mean DBH, height, and number of trees for each class.Here, we used trembling aspen functions for hardwood species, jack pine functions for all pine species and pine-mixture, black spruce functions for all spruce and fir species, and mixed-stand function for mixed softwoods/hardwoods.The Canada-wide biomass equations calculate the oven-dry biomass of stem, bark, and crown of a single tree for 18 Canadian tree species (Evert, 1985), with a given DBH and height.Using the information of tree-size distribution and the biomass equations, we first calculated each biomass component (stem, bark, and crown), added these components to get total aboveground biomass, and then multiplied with the number of trees in each tree-size distribution class to get the total tree biomass per hectare, B t , in t ha -1 .The aboveground oven-dry biomass of clustered high shrub and regenerating trees at a site, B s in t ha -1 , was calculated by 12 () 25 where (b s1 + b s2 )/2 is the mean green biomass of the two clusters of shrub sampled or that of regenerating trees (kg per cluster or regenerating tree), N c is the number of clusters of shrub or regenerating trees in the 3.99 m-radius plot, and R s is the ratio of oven-dry to fresh biomass for shrub or the regenerating trees.The aboveground oven-dry biomass of graminoids, herbs and shrubs in the lower layers, B g in t ha -1 , was calculated by 12345 () 10 5 where b g1 , b g2 , b g3 , b g4 , and b g5 are, respectively, the aboveground green biomass of the graminoids/herbs and shrubs mixture for each of the five plots in a site (kg m -2 ), and R g is the ratio of oven-dry to fresh biomass for the mixture of graminoids/herbs and shrubs.The oven-dry biomass of lichen, B l in t ha -1 , was calculated by 1234512 () () 1000 52 where P l1 , P l2 , P l3 , P l4 , and P l5 are, respectively, the percentage of lich ground cover at the 5 plots, b l1 and b l2 are, respectively, the fresh lichen biomass of each of the two 10-cm by 10-cm samples collected from one of the plots (kg m -2 ), and R l is the ratio of oven-dry to fresh biomass for lichen.The oven-dry biomass of live moss, B m in t ha -1 , was calculated by 1234512 () () 1000 52 where P m1 , P m2 , P m3 , P m4 , and P m5 are, respectively, the percentage of moss ground cover at the 5 plots, b m1 and b m2 are, respectively, the aboveground green biomass of each of the two 1-cm depth, 10 cm by 10 cm square moss samples collected from one of the plots (kg m -2 ), D m is the depth of alive moss in cm, and R m is the ratio of oven-dry to green biomass for moss.The aboveground biomass for a site, B a , was thus the summation of all these components where proper.
The calculation of foliage biomass, B f in unit g m -2 , was similar to that of aboveground biomass, except that only the foliage component of vascular plants was included.Elvidge, 1996;Chen et al., 2010a).Furthermore, clouds and forested areas were removed and masked, and the mosaic was clipped based on the Porcupine caribou habitat boundary.

Remote sensing data sources and processing
The JERS-1/SAR datasets were extracted from the North America JERS mosaics (acquired in the summer of 1998) to cover the Porcupine caribou habitat (Kyle McDonald, JPL, personal communication).In addition, data of 1:50000 DEM covering the habitat were also obtained for orthorectifying the JERS mosaic.The North America JERS summer mosaics were not precisely geo-referenced, because topographic distortions were not removed during the mosaic process due to the lack of adequate DEM (Sheng & Alsdorf, 2005).Visual inspection showed that the offset between the JERS mosaic and the Landsat images may be 300 m ~ 5000 m.The positioning error was expected to be much larger in mountainous regions.
In this study, we employed a SAR image simulation method from DEM data to correct topographic distortions (Sheng & Alsdof, 2005).Briefly speaking, this method includes four steps: (1) simulating a SAR image in an azimuth-range projection from the DEM according to imaging geometry of real SAR image; (2) collecting ground control points that tie the uncorrected SAR image to the simulated SAR image; (3) warping the real SAR image to the simulated SAR image using a polynomial function fitted from the ground control points; and (4) projecting the warped real SAR image back to the DEM map coordinate system.The method requires three types of inputs: individual scene of SAR imagery, DEM data, and SAR imaging geometry parameters (i.e.sensor altitude H, minimum look angle θ, the orbital azimuth angle β).Since the year-day file, which contains the date of specific JERS image's acquisition, was delivered with the JERS mosaics data, we could extract individual path image in the JERS mosaic.The path image was used as an input in place of individual scene of JERS.In addition, since the imaging geometry information of the individual path in the JERS mosaic was unavailable, general JERS imaging geometry parameters were used in the SAR simulation method.For the purpose of image matching for ground control point selection, a correct geometry was preferred but is not necessary (Sheng and Alsdof, 2005;Linder and Meuser;1993).In this study, we used H = 568000 meter, θ =35 degree, and β =190 degree in SAR image simulation.Finally, the Digital Number (DN) values of JERS mosaic could be converted to backscatter coefficients σ (in db format) by the equation: The co-registered and geo-referenced Landsat TM/ETM+ and JERS-1 mosaics were then compared with measurements of aboveground and foliage biomass, with their best-fit relationships being applied back to the mosaics to produce maps of aboveground and foliage biomass over the Porcupine caribou habitat, as outlined in Fig. 4.

Mid-growing season
Landsat images

Mid-growing season Landsat mosaic
Baseline foliage biomass map

Aboveground biomass field measurements
Fig. 4. Flow chart describing the procedures of using field measurements and remote sensing data to produce aboveground and foliage biomass baseline maps.

Measured aboveground and foliage biomass over the Porcupine caribou habitat
Table 1 summarizes the mean value, standard deviation, and range of aboveground biomass measured at sites within and around the Porcupine caribou winter and summer range along the Dempster Highway, Yukon, during the summer of 2004.Average values of measured aboveground biomass of sparsely treed woodland, low-high shrub lands, and mixed graminoids-dwarf shrub-herb lands were, respectively, 57.3, 11.1, and 2.3 t ha -1 .Within each vegetation type, the ranges of measured aboveground biomass were very large.The standard deviations of measured aboveground biomass among sites were often larger than their corresponding mean values, especially for low-high shrubs and mixed graminoidsdwarf shrub-herb, and when all types of vegetation were considered.The measured aboveground biomass ranged from 10 to 100 t ha -1 for sparsely treed woodlands, from 1 to 100 t ha -1 for the low-high shrub sites, from 0.5 to 10 t ha -1 for mixed graminoids-dwarf shrub-herb sites.These measurements indicate that there are significant overlaps in the ranges of aboveground biomass between sparsely treed woodlands and low-high shrub sites, and between low-high shrub sites and mixed graminoids-dwarf shrub-herb sites.Similarly, the foliage biomass values measured at sites within a specific dominant vegetation type also varied significantly (Table 2).For example, foliage biomass ranges from 95.3 to 198.4 g m -2 for low-high shrub sites, 37.9-92.3g m -2 for mixed graminoids-dwarf shrub-herb sites, 63.1-106.1 g m -2 for coastal plain tussock sites, and 0.0-20.0g m -2 for hilltop rock lichen sites.Consequently, assigning aboveground or foliage biomass value to a site according to its vegetation type can result in substantial error (Gould et al., 2003;Walker et al., 2003).

Relationships between aboveground biomass and remote sensing indices
In this study, we investigated the applicability of both optical and radar data as well as their combinations.To obtain the best regression model for estimating aboveground biomass, robust multiple regressions were conducted between field-measured aboveground biomass and various remote sensing-derived variables such as TM spectral reflectance, vegetation indices, and JESR-1/SAR backscatter coefficients.We used 3 × 3 pixels (i.e., 90 m by 90 m) averaged value in place of single pixel value in order to reduce the effect of erroneous spectral features, e.g., features of adjacent pixels may have been assigned to some field plots of the data due to errors in image registration and the location of sample plots.The Landsat images were re-sampled to 100 m resolution for matching the resolution of the North America JERS summer mosaic.
For the sites along the Dempster Highway, we found that strong correlations exist between ln(B a ) and remote sensing signals (Table 3).When all types were mixed, the strongest correlation was found against the L-band JERS-1/SAR backscatter, followed by the Landat B4/B5 (Fig. 5).The Landsat bands 3 and 5 show strong negative relationships for the sparse woodlands and all types mixed, but not for shrub and graminoids lands.Many vegetation indices have been developed and applied for estimating aboveground biomass (Anderson & Hanson, 1992;Anderson et al., 1993;Mutanga & Skidmore, 2004;Lu 2005).In Table 3, we examined 4 vegetation indices: ratio of Landsat B4/B5, simple ratio (SR = B4/B3), normalized differential vegetation index (NDVI = (B4 -B3)/(B4 + B3)), shortwave vegetation index (SWVI = (B4 -B5)/(B4 + B5)).In general, vegetation indices can partially reduce the impacts on reflectance caused by environmental conditions and shadows, thus improve correlation between AGB and vegetation indices, especially in those sites with complex vegetation stand structures (Lu et al., 2004).However as shown in Table 3, not all vegetation indices are significantly correlated with aboveground biomass.Consequently, various degrees of success had been obtained in estimating aboveground biomass using Landsat vegetation indices (Sader et al., 1989;Lee & Nakane, 1996;Nelson et al., 2000;Steininger, 2000;Foody et al., 2003;Phua & Saito, 2003;Zheng et al., 2004).For example, Nelson et al. (2000) found that aboveground biomass cannot be reliably estimated using Landsat data without the inclusion of secondary forest age.Steininger (2000) explored the ability of Landsat data for estimating aboveground biomass of tropical secondary forests and found that saturation was a problem for advanced successional forests.
Similarly, different degrees of success had been obtained in previous studies of using radar data for estimating aboveground biomass, with longer-wavelength L-band and Pband SAR data proven to be more valuable (Sader, 1987;Le Toan et al., 1992;Rauste et al., 1994;Ranson et al., 1997;Luckman et al., 1997;Kurvonen et al., 1999;Kuplich et al., 2000;Tsolmon et al., 2002;Sun et al., 2002;Castel et al., 2002;Santos et al., 2002).For example, Kuplich et al. (2000) used JERS-1/SAR data for aboveground biomass estimation of regenerating forests.Sun et al. (2002) found that multi-polarization L-band SAR data were useful for estimating aboveground biomass of forest stands in mountainous areas.Castel et al. (2002) identified the significant relationships between the backscatter coefficient of JERS-1/SAR data and the stand biomass of a pine plantation.Santos et al. (2002) used JERS-1/SAR data to analyse the relationships between backscatter signals and biomass of forest and savanna formations.The significant correlation between aboveground biomass and JERS-1/SAR backscatter coefficient, as shown in Table 3 and Fig. 4, indicates that longer-wavelength L-band SAR data are also valuable in the Arctic.Nevertheless, the saturation problem is also common in estimating aboveground biomass using radar data (Luckman et al., 1997;Balzter, 2001, Lucas et al., 2004;Kasischke et al., 2004).For example, Luckman et al. (1997) found that the longer-wavelength L-band SAR image was more suitable to discriminate different levels of forest biomass up to a certain threshold than shorter-wavelength C-band SAR data.
To take advantages of the ability of Landsat vegetation indices and JERS-1/SAR backscatter coefficient, we used both data for estimating aboveground biomass.The best fit relationship between aboveground biomass and JERS-1/SAR backscatter coefficient σ as well as Landsat B4/B5 for all types mixed in the Dempster Highway study area is given by: ln( ) 2.3759(B4/B5)+0.5542+4.0948 with a coefficient of determination r 2 = 0.72, and standard estimation error (SEE) = 0.78.Because the logarithmic equations could introduce a systematic bias when used for back calculating biomass, it has now become fairly widely recognized that a correction factor is necessary to counteract this bias (Sprugel, 1983).The correction factor (CF) can be calculated by using the formula: For equation ( 7), the CF = 1.35.The 1:1 comparison of estimated aboveground biomass using CF-corrected equation ( 7) and the measured values for the sites along the Dempster Highway is shown in Fig. 6, with the slope = 0.95, r2 = 0.72, and SEE = 13.5 t h -1 over measured aboveground biomass range from 0.4 to 94.5 t h -1 .These results agreed well with the finding of Moghaddam et al. (2002) that the estimation accuracy of forest biomass was significantly improved when radar and optical data were used in combination, compared to estimates using a single data type alone.

Validation of aboveground biomass relationship with remote sensing indices
In a previous study (Chen et al., 2009a), we validated equation ( 7) using aboveground biomass measurements at 33 sites around Yellowknife, Northwest Territories, and Lupin Gold Mine, Nunavut Territories.The 1:1 comparison of estimated aboveground biomass and the measurements for the Yellowknife and Lupin Gold Mine study area indicates that the equation holds up very well, with r 2 = 0.81, slope =1.17, and SEE = 9.67 t h -1 over measured aboveground biomass range from 0.9 to 103.3 t h -1 .In this study, we further validated the relationship using measurements collected over the same study area at later dates.We used aboveground biomass measured at 21 sites along the Dempster Highway during the summer of 2006 in the Ivvavik National Park during the summer of 2008, where foliage biomass was the main measurement target, to validate the above relationship.Fig. 7 shows the validation result, with the slope = 0.91, r 2 = 0.90, and SEE = 1.5 t h -1 over measured aboveground biomass range from 0 to 18.16 t h -1 .These validation results suggest that the aboveground biomass can be reliably estimated using Landsat B4/B5 and JERS-1/SAR backscatter coefficient.

Relationships between foliage biomass and remote sensing indices
As shown in Table 4, the Landsat-based SR can explain 81% variations in measured foliage biomass within the Porcupine caribou habitat.The explanation power of other Landsatbased indices in the descending order is 76% for the ratio of B4/B5, 71% for SWVI, and 67% for NDVI.Single band reflectance of Landsat has lower power of explanation.In comparison with results for aboveground biomass shown in Table 3, Landsat-based vegetation indices have a much improved power of explanation for foliage biomass (Table 4).This is in good agreement with the fact that optical sensors mainly capture canopy information, thus the optical sensor data are more suitable for estimation of canopy parameters such as foliage biomass than aboveground biomass, as demon stared by previous studies (Franklin & Hiernaux, 1991;Hall et al., 1995;Chen & Cihlar 1996;Turner et al., 1999;Brown et al., 2000;Chen et al., 2002;Wylie et al., 2002;Phua & Saito, 2003;Laidler & Treitz, 2003;Lu, 2004).The best fit relationship between foliage biomass (B f , g m -2 ) and Landsat-based simple ratio (SR) over the Porcupine caribou habitat is given as follows 16.62 -1.1906 with r 2 = 0.81, SEE = 20.6 g m -2 , F= 81, P = 2.7×10 -8 , and n = 21 (Fig. 8).Because we had only a relative small sample size of foliage biomass over the Porcupine caribou habitat, we didn't leave a fraction of the foliage biomass measurement points as validation.Nevertheless, we did find a similar relationship between Landsat-based simple ration and foliage biomass for the Bathurst caribou habitat located in Northwest Territory, Nunavut Territory, and northern Saskatchewan (Chen et al., 2011), with r 2 = 0.86, SEE = 26.3g m -2 , F= 158, P = 2.6×10 -12 , and n = 27.

Baseline maps of aboveground and foliage biomass
Applying equations ( 7) and ( 8) to the co-registered and geo-referenced Landsat and JERS-1/SAR mosaics data over the Porcupine caribou habitat, we produced aboveground biomass for the Porcupine caribou habitat.Fig. 9 shows aboveground biomass distribution in circa 2000 over the Porcupine caribou habitat.The amount of aboveground biomass was less than 5 t ha -1 for most area of the Porcupine caribou calving ground located in northern coastal area in both Alaska and Yukon.Along the coastal line and at high mountain ridges, the aboveground biomass further decreased to less than 2 t ha -1 .For the summer and winter ranges of the Porcupine caribou habitat to the south, the values of aboveground biomass generally increased, with a large percentage of the areas having aboveground biomass in the range of 5-20 t ha -1 , as well as a significant fraction of the area of 2-5 t ha -1 aboveground biomass.The areas of aboveground biomass 2-5 t ha -1 appeared to be dominated by graminoids and herbs, while those of 5-20 t ha -1 aboveground biomass were mainly shrub land.The Porcupine summer and winter ranges also had a small fraction of area that had more than 20 t ha -1 aboveground biomass, likely associated with treed woodland.Areas of 0-2 t ha -1 aboveground biomass also occur in the summer and winter ranges of the Porcupine caribou habitat, mostly happened at high mountain ridges.The information on foliage biomass can be more useful to local caribou management boards and researchers because it is directly related forage availability (Russell et al., 1993;Russell & McNeil, 2002;Russell et al., 2002).Using equation ( 9), we developed circa 2000 foliage biomass distribution over the Porcupine caribou habitat (Fig. 10).10.We also blocked out water and cloudcover areas in Fig. 10.As shown in Fig. 10, most of the northern coastal areas in Alaska and Yukon, over which the calving ground of the Porcupine caribou habitat is located, had a quite high foliage biomass from 50 to > 100 g m -2 .This is a contrast to the distribution of the aboveground biomass.Over these same coastal areas the aboveground biomass was quite low.Both aboveground biomass and foliage biomass were found to be lower at high mountain ridges, and coastal beaches.The foliage biomass map reveals that on average the amount of seasonal peak foliage biomass in the calving ground of the Porcupine caribou herd was similar to that in the summer range, and was even higher if only the concentrated calving ground over the Northern Alaska and Yukon coastal plain is concerned (Table 5).To the contrast, the average seasonal peak foliage biomass in the calving ground of the Bathurst caribou herd was much lower than that in the summer range during the same time period (Chen et al., 2011).This result agree well with the ground survey by Griffith et al. (2001), which suggested that on June 14, during 1998-1999, the total available forage was 63 g m -2 for the Bathurst caribou calving ground, in comparison with 460 g m -2 for the Porcupine caribou calving ground.Of the 63 g m -2 forage over the Bathurst caribou calving ground, 33 g m -2 was lichen and moss, 14 g m -2 was stand dead, and only 16 g m -2 was vascular biomass.On the contrast for the Porcupine caribou calving ground, 50 g m -2 was lichen and moss, 250 g m -2 was stand dead, and 160 g m -2 was vascular biomass.The live biomass over the Porcupine calving ground was 10 time that over the Bathurst calving ground.In addition, the high value of stand dead over the Porcupine calving ground also suggest the high foliage biomass value over previous years.The difference in calving ground foliage biomass collaborates well with the fact that cows of the Bathurst herd leave calving ground soon after giving birth while those of Porcupine herd stay for a much longer period (Griffith et al., 2001) The difference in calving ground forage availability, in turn, results in difference in caribou biological measures (Table 5).For example, the body weights of Porcupine cows are about 20% higher than that of Bathurst herd (Griffith et al., 2001).The Bathurst cow milk production was 1.09 -1.79 liter per day, against 2.02 liter per day by an average Porcupine caribou cow, with the corresponding calve growth rate being 150 -407 gram per day for the Bathurst herd against 493 gram per day for that of the Porcupine herd (Table 5).

Conclusions
Aboveground biomass was measured at 43 sites in the summer of 2004 along the Dempster Highway, which goes through the winter and summer ranges of the Porcupine caribou habitat.The measured aboveground biomass ranged 10-100 t ha -1 for sparsely forested woodlands, 1-100 t ha -1 for the low-high shrub sites, and 0.5-10 t ha -1 for mixed graminoidsdwarf shrub-herb sites.
Foliage biomass was measured at 10 non-forested sites along the Dempster Highway in the summer of 2006, and again in the summer of 2008 at 11 non-forested sites in the Ivvavik National Park located at northern tip of Yukon, which overlaps with the calving ground and summer range of the Porcupine caribou herd.The measured foliage biomass ranged 0.95-2 t ha -1 for low-high shrub sites, 0.38-0.92t ha -1 for mixed graminoids-dwarf shrub-herb sites, 0.63-1.06t ha -1 for coastal tussock sites, and < 0.2 t ha -1 for hill-top rock lichen sites.When all data points were pooled together, the best relationship between aboveground biomass and remote sensing data was found to be with the combination of JERS-1 backscatter and Landsat B4/B5, with r 2 = 0.72.Validation using aboveground biomass measurements at foliage biomass measurement sites gives similar result, with the slope = 0.91, r 2 = 0.90, and SEE = 1.5 t h -1 over measured aboveground biomass range from 0 to 18.16 t h -1 .Similar validation results were obtained using aboveground biomass measurements at 33 sites around Yellowknife, Northwest Territories, and Lupin Gold Mine, Nunavut Territories in a previous study (Chen et al., 2009a), with r 2 = 0.81, slope =1.17, and SEE = 9.67 t h -1 over measured aboveground biomass range from 0.9 to 103.3 t h -1 .For the foliage biomass, the Landsat-based simple ration gives the best fit, with r 2 = 0.81, SEE = 20.6 g m -2 , F= 81, P = 2.7×10 -8 , and n = 21.Appling these relationships to the mosaic of Landsat and JERS-1 images covering the entire Porcupine caribou habitat, we produced baseline maps of aboveground biomass and foliage biomass for the Porcupine caribou habitat.The foliage biomass map reveals that on average the amount of seasonal peak foliage biomass in the calving ground of the Porcupine caribou herd was similar to that in the summer range, and was even higher if only the concentrated calving ground over the Northern Alaska and Yukon coastal plain is concerned.To the contrast, the average seasonal peak foliage biomass in the calving ground of the Bathurst caribou herd was much lower than that in the summer range during the same time period.The difference in calving ground foliage biomass collaborates well with the fact that cows of the Bathurst herd leave calving ground soon after giving birth while those of Porcupine herd stay for a much longer period, which in turn partially explain why the Porcupine caribou calves can have several times higher growth rate and the body weights of Porcupine cows are about 20% higher than that of Bathurst herd.
It should be emphasized, however, that assessment of the impacts of climate change on the Porcupine caribou habitat requires integration of various information sources from remote sensing products, climate records, and other relevant tempo-spatial data.Furthermore, there are many other factors besides the habitat, such as harvest, predators, diseases/parasites, industrial development, extreme weather events, climate change, and pollution, may influence the abundance of a caribou herd.Therefore, this work is just one of the first steps towards informed and proper decision-making that balances the needs of caribou habitat protection and industrial development under a global change environment.

Acknowledgment
This study is financially supported by grants from the Canadian IPY program through a project entitled "Climate Change Impacts on Canadian Arctic Tundra Ecosystems (CiCAT): Interdisciplinary and Multi-scale Assessments", Canadian Space Agency's Government Related Initiatives Program (GRIP) through a project entitled "ParkSPACE: Towards an Operational Satellite-based System for Monitoring Ecological Integrity of Arctic National Parks", and from the Earth Sciences Sector, NRCan's "Climate Change Geosciences

Fig. 1 .
Fig. 1.Locations of the Porcupine caribou calving ground, summer range, and winter range, as well as that of the aboveground and foliage biomass measurement sites along the Dempster Highway transect during the summers of 2004 and 2006, and in the Ivvavik National Park during the summer of 2008.

Fig. 2 .
Fig. 2. Plot design for measuring aboveground and foliage biomass at non-forested sites.

Fig. 3 .
Fig. 3. Photograph showing biomass sampling at a coastal plain site in the Ivvavik National Park, July 25, 2008.Front are 3 northern students from Aklavik, NWT (left to right: Alexander Gordon, Jayneta Pascal & Kayla Arey), and at the background Weiroing Chen of CCRS was sampling root biomass (photo by Wenjun Chen).

Fig. 6 .
Fig. 6.A 1:1 comparison between estimated and measured aboveground biomass values for the calibration sites along the Dempster Highway transect measured during the summer of 2004.For clarity over the low biomass range, the results are shown on the log-log scale.
Fig. 7.A 1:1 comparison between estimated and measured aboveground biomass values for the validation sites along the Dempster Highway transect measured during the summer of 2006 and in the Ivvavik National Park during the summer of 2008.For clarity over the low biomass range, the results are shown on the log-log scale.

Fig. 8 .
Fig. 8. Relationship between Landsat simple ratio and foliage biomass measured at sites along the Dempster Highway during the summer of 2006 and in the Ivvavik National Park during the summer of 2008.

Fig. 9 .
Fig. 9. Aboveground biomass distribution over the Porcupine caribou habitat circa 2000.The 3-d effect was generated using the DEM data.

Fig. 10 .
Fig. 10.Foliage biomass distribution over the Porcupine caribou habitat cirac 2000.The 3-d effect was generated using the DEM data.Areas covered by water, forest, and cloud were blocked out.

Fig. 10
Fig.10shows the circa 2000 foliage biomass distribution of non-forest land areas in the Porcupine caribou habitat.Because of lacking foliage biomass measurements over forested land, foliage biomass was not estimated over forest areas in the Porcupine caribou habitat and areas covered forests were blocked out in Fig.10.We also blocked out water and cloudcover areas in Fig.10.As shown in Fig.10, most of the northern coastal areas in Alaska and Yukon, over which the calving ground of the Porcupine caribou habitat is located, had a quite high foliage biomass from 50 to > 100 g m -2 .This is a contrast to the distribution of the aboveground biomass.Over these same coastal areas the aboveground biomass was quite low.Both aboveground biomass and foliage biomass were found to be lower at high mountain ridges, and coastal beaches.The foliage biomass map reveals that on average the amount of seasonal peak foliage biomass in the calving ground of the Porcupine caribou herd was similar to that in the summer range, and was even higher if only the concentrated calving ground over the Northern Alaska and Yukon coastal plain is concerned (Table5).To the contrast, the average seasonal peak foliage biomass in the calving ground of the Bathurst caribou herd was much lower than that in the summer range during the same time period(Chen et al., 2011).This result agree well with the ground survey byGriffith et al. (2001), which suggested that on June 14, during 1998-1999, the total available forage was 63 g m -2 for the Mapping Aboveground and Foliage Biomass Over the Porcupine Caribou Habitat in Northern Yukon and Alaska Using Landsat and JERS-1/SAR Data Nearly clear-sky Landsat TM/ETM+ level 1G orthoimagery was downloaded from the United Stated Geological Survey (USGS) website (http://earthexplorer.usgs.gov)and from the Centre for Topographic Information (CTI) of Natural Resources Canada, available through Geogratis (http://geogratis.cgdi.gc.ca/).23 scenes are needed to cover the entire Porcupine caribou habitat.Most of them were acquired within the middle growing season (July 10 to August 15) during 1999 -2003.Only Bands 3 (0.63-0.69 μm), 4 (0.75-0.90 μm), and 5 (1.55-1.75μm) were used in the study, because Bands 1-3 are highly correlated, as are Bands 5 and 7. Prior to further analysis, all scenes, if necessary, were re-projected to Lambert Conformal Conic (LCC) projection with 95 o W, 49 o N as the true origin and 49 o N and 77 o N as standard parallels.Surface reflectance was derived by radiometrically normalizing each scene to 250 m resolution clear sky MODIS imagery, which was a 10-day composite acquired during July 21-31, 2001 with matching bands calibrated to a strip of atmospherically corrected Landsat TM/ETM+ images.A Landsat mosaic of surface reflectance was then generated with further normalizations for individual scenes if substantial discrepancy exists.All radiometric normalization and calibration equations were developed using a Scattergram Controlled Regression method(Elvidge et al., 1995; Yuan &

Table 1 .
Mean value, standard deviation, and range of aboveground biomass measured at sites within and around the Porcupine caribou winter and summer range along the Dempster Highway, Yukon, during the summer of 2004.

Table 2 .
Mean value, standard deviation, and range of foliage biomass measured at sites within and around the Porcupine caribou winter and summer ranges along the Dempster Highway, Yukon during the summer of 2006, as well as at sites within and around the Porcupine caribou calving ground and summer range inside the Ivvavik National Park, Yukon during the summer of 2008.Also shown are statistics for the corresponding aboveground biomass measurements.

Table 3
. Correlation coefficient (r) and coefficient of determination (r 2 ) between ln(B a

Table 4 .
Correlation coefficients (r), coefficient of determination (r 2 ), and standard estimation error (SEE, in g m -2 ) between foliage biomass (B f

Table 5 .
(Chen et al., 2011)sured and area-averaged foliage biomass values between Porcupine caribou habitat and Bathurst caribou habitat.The area-averaged values were calculated from the circa 2000 baseline foliage biomass maps.Field measurement of foliage biomass over the Bathurst caribou habitat was conducted during the summer of 2005(Chen et al., 2011).Also included in the comparison are cow milk production, calve growth rate, and cow body weight.