Flux measurements at sites of mixed hardwood and black spruce stands from an area (C4) of the Caribou-Poker Creek Research Watershed (CPCRW), interior Alaska, in the summer seasons of 1998, 1999, and 2000 are used to estimate the fluxes of CO2, CH4, and N2O before and after forest fires. The FROSTFIRE burning experiment was executed in typical boreal forest from 8 to 15 July 1999. The forest fire significantly decreased soil CO2 and N2O emissions, by at most 50%. On the contrary, CH4 flux from the soil increased from 7 to 142%, suggesting that the forest fire plays a role in accelerative thawing of the frozen soil, and subsequently the release of CH4 from permafrost. Most of the CH4 was oxidized in the soil after the fire; however, some was released from the soil when the permafrost maximally thawed at the end of August 1999 and September 2000. Relationships between the fluxes of trace gases and soil temperature before and after the fire showed good exponential correlations, indicating that soil temperature was one of the factors determining the fluxes of trace gases on boreal forest soils. Also, the higher soil temperature after the fire may be led to the enhanced diffusion of CO2, CH4 and N2O by microbial activity between the atmosphere and the forest soils, and to the increased fluxes of trace gases in burned black spruce stand soils. In order to understand the roles of moss and lichen mats on the black spruce stands, the net respiration by mosses and lichens was estimated with light and dark chamber measurements. Net respiration corresponds to 42 to 58% of the total soil respiration before fire. Therefore, the net respiration by moss and lichen layers was responsible for one-half of total soil CO2 emissions. The maximum regional net respiration rate by moss and lichen mats on black spruce forest floors of central Alaska was 0.018 ± 0.009 GtC/yr, an important source of atmospheric CO2 in boreal forests. After the prescribed burn, soil respiration was attributable only to respiration by roots and microbes. The microbial respiration estimated after the fire is almost three times as high as that the calculated respiration before the fire. This finding indicates the post-fire condition may stimulate microbial respiration because of higher nutrients and substrates in remnant soils and enhanced soil temperature. The microbial respiration can be estimated 14.7 tC/ha in burned black spruce stands over a decade after the fire, suggesting burned black spruce forests in central Alaska are a crucial source of atmospheric CO2.
 Boreal forests account for about one-third of the carbon sequestered in terrestrial ecosystems. These are located in the high latitudes, in areas including Alaska, Canada, Russia, and Northern Europe. Higher latitude ecosystems are particularly vulnerable to climate change due to the large amount of carbon in northern latitude soils and the predominance of permafrost. Northern boreal forests represent approximately 35% of the world's forests; including permafrost. Northern boreal forests contain approximately 66% of the world's forest soil carbon pools [Van Cleve et al., 1983; Oechel and Vourlitis, 1997; Billings, 1997; Kasischke and Stocks, 2000]. Because boreal forests absorb atmospheric carbon dioxide and slowly decompose the litter, fibric and humic substances, the ecosystems are known as carbon sinks [Schlesinger, 1997; Fan et al., 1998].
 Forest fire is a major disturbance in boreal forests, with its occurrence closely coupled to climate patterns. Therefore, changes in climate will result in change to the fire regime as well. Although boreal forests are presently one of the major terrestrial carbon pools, shifts in the fire regime and ecosystem distribution in high latitudes associated with climate change are likely to result in significant increases in atmospheric concentrations of carbon dioxide and other greenhouse gases [Kasischke et al., 1995; Kasischke, 2000; Kasischke et al., 2000a, 2000b; Richter et al., 2000]. As boreal forests emit higher concentrations of carbon to the atmosphere immediately after the fire, forest fires in the northern stands are well known as carbon sources [Seiler and Crutzen, 1980; Crutzen and Andreae, 1990; Levine, 1991; Kasischke et al., 2000b, 2000c]. Hansen et al.  showed that, based on the average temperature over the past 30 years, the most significant areas of warming coincide with the region occupied by the boreal forest. Given the close linkage between fire occurrence and climate, there should be little surprise that over the past two decades, there has been a significant increase (almost threefold) in the annual area burned in the North American boreal forest [Kasischke and Stocks, 2000].
 Most previous researchers have estimated the amount of CO2 released by fire in boreal forests [Kasischke et al., 1995, 2000a, 2000b, 2000c; Kasischke, 2000; Levine and Cofer, 2000] and soil respiration rate after fire [O'Neill et al., 1997; Kasischke et al., 2000b; Richter et al., 2000]. However, there are few reports on the exchange of CO2, CH4 and N2O before and after boreal forest fire. The landscape-scale prescribed research of burn in the boreal forest of interior Alaska, FROSTFIRE, was carried out from July 8 to 15, 1999. Within the 2200-acre perimeter, the fire consumed mostly black spruce stands (900 acres). This burning experiment differs from previous experimental fires in the boreal forest because it is in terrain dominated by permafrost, focuses on the large-scale ecological consequences of fire, and takes place in a Long Term Ecological Research (LTER) site, enabling long-term, experimentally-controlled research. The objectives of this study are to clarify the effects of forest fire on the fluxes of CO2, CH4 and N2O in black spruce stands, to understand the main factor determining the fluxes of trace gases in unburned and burned stands, to evaluate the net respiration rate by aboveground vegetation (e.g., mosses and lichens) on the soil surface of black spruce stands, and finally, to imply global budgets of trace gases after the fire.
2. Sampling and Analytical Methods
2.1. Sampling Locations
 In Figure 1, sampling locations are shown in Caribou-Poker Creeks Research Watershed (CPCRW). Annual average air temperature was −3.3°C, and the average temperature in July was 16.4°C, which was the maximum in this watershed. Annual average precipitation was 269 mm, and the average precipitation of 47 mm in August was the highest in this area. In a mixed hardwood stand, two observation sites were placed in the lower creek (site ‘A’) and in the upper creek (site ‘B’). Sites A and B each had three sampling points, which were located on a south-facing slope. The black spruce station (‘BS’ site) was set on the opposite side of the creek from site B, facing north, and has three sampling points. Sampling sites A and B are located in a mixed deciduous forest with the dominant vegetation of highly productive aspen (Populus tremulides Michx.) and paper birch (Betulus papyrifera Marsh). Site BS is in coniferous forest, which consists of low productivity moss-dominated black spruce (Picea mariana (Mill.) B.S.P.). The forest floor is covered with the moss and lichen, Ptilium crista-castrensis and Pleurozium schreberii, respectively. The discontinuous permafrost is widely distributed at 50 cm beneath the black spruce forest floor. Site BS, burned in 1996, is located on the outside of the CPCRW boundary, and the vegetation is similar to black spruce stand in CPCRW.
 The soil profile in mixed hardwood forest showed fallen leaves and branches within 5 cm of the surface, a humic layer (5–20 cm), and living hardwood roots and clayey soils below. The profile in black spruce indicated deeper moss and lichen layers (0–25 cm), a humic layer (25–50 cm), and the permafrost below.
2.2. Flux Measurement
 Flux observations for this study were performed during three periods, 9 July to 14 August 1998, 1 July to 31 August 1999, and 26 July to 26 September 2000. The static chamber method was employed to estimate the fluxes of CO2, CH4 and N2O at each stand soil surface. A stainless collar was used for supporting the chamber and sealing it against leakage. Two kinds of chambers (30 cm diameter, 10 cm high) were used in the summer season of 1998. One is made with transparent material (acrylic) and the other with non-transparent material (stainless steel). Hereafter, these are called “light” and “dark” chambers, respectively. After 1999, a smaller chamber (20 cm diameter, 7 cm high) was also used for flux measurements. The chamber was mounted on the soil surface of mixed hardwood stands, and on moss and lichen floors of black spruce forests, and the sampling was carried out using a 50 mL disposable syringe, at a sampling rate of about 30 ml/min. The air samples from the chamber were collected five times at 5 to 10 min intervals, and transferred into a 100 mL aluminum foil-sampling bag (GL Science, Japan). The flux was calculated from the gradient of the relationship between the concentration variations of a gas and sampling time. The correlation coefficient was above 0.99. The soil temperature was measured using a portable thermometer at 5 cm depths at sites A and B, and at 20 cm depths under the moss and lichen mats at the BS site.
3. Analytical Method
 CO2 concentration from the sampling bag was measured within 10 hours after the sampling. One mL of air sample was taken by using a syringe and injected into a thermal conductivity detector gas chromatograph (TCD-GC, Shimadzu Co. Ltd., Japan) with a column packed with Porapack Q (80/100 mesh). The measurements were calibrated with a series of standard gases containing 338 ± 7, 491 ± 10 and 5000 ± 100 ppmv of CO2 with WMO scale. The analytical error was usually less than 1%. In situ measurements of CO2 concentrations were also conducted with portable TCD-GC in summer of 1998.
 One mL of air sample was injected into a flame ionization detector (FID-GC, GC-14B, Shimadzu Co. Ltd., Japan) with a column packed with Molecular Sieve 13X (30/50 mesh). The concentration was calibrated against the three kinds of standard containing 0.79 ± 0.016, 2.43 ± 0.048 and 5.01 ± 0.089 ppmv of CH4. The analytical error was less than 3%.
 One mL air sample was taken out from the sampling foil and injected into a 63Ni electron capture detector gas chromatograph (ECD, GC-14B, Shimadzu Co. Ltd., Japan) with a column packed with Porasil (80/100 mesh). The calibration was done with series of standard gases containing 302 ± 6, 784 ± 8 and 1308 ± 15 ppbv of N2O. The analytical error was less than 2%.
4. Results and Discussion
 Fluxes of CO2, CH4 and N2O and soil temperature observed in mixed hardwood and black spruce stands in the summer seasons of 1998, 1999, and 2000 before and after the prescribed fire, and previously reported results in central Alaska, are given in Table 1.
Table 1. Average Fluxes of CO2, CH4 and N2O and Soil Temperature Observed in Mixed Hardwood and Black Spruce Stands in Summer Seasons of 1998, 1999 and 2000, and Reported Values in the Literatures for Interior Alaska
Vegetation type (Site, observation year)
CO2 Flux (mmol/m2/h)
CH4 Flux (μmol/m2/day)
N2O Flux (μmol/m2/h)
N.D. indicates no data.
The fluxes measured in moss area in birch and aspen forests.
The flux observation after the removal of moss and lichen mats.
Black spruce forests were burned in 1996 and flux measurements carried out in 1998.
L and D denote light and dark chambers, respectively.
 Because of the thicker coverage of moss and lichen on the floor in 1998 and 1999, the fluxes of CO2, CH4 and N2O with light and dark chambers were estimated in these stands. Particularly, the flux chamber measurements were carried out in order to evaluate the net respiration rate by aboveground vegetation that are mainly mosses and lichens as well as shrubs on black spruce floors. Here, we examined the measurements only in moss and lichens. We describe the net respiration rate by moss and lichen layers in section 5.
 Based on the t-test, there were significant differences in average fluxes of CO2, CH4 and N2O estimated between light and dark chambers (p < 0.05). The CO2 fluxes were similar to values of Kakukawa  and Valentine and Boone  observed near the observation sites at black spruce stand in CPCRW, and were one-half of values measured by Schlentner and Van Cleve  and Gordon et al.  in interior Alaska, and by Weber  in Ontario jack pine ecosystems. After the removal of moss and lichen layers from the forest floor, we found using a light chamber the average CO2 flux was 0.4 ± 0.1 mmol/m2/h (using a light chamber), and soil temperature was 0.7°C at a soil depth of 20 cm, which was lower than Kakukawa's data in a wide soil temperature range of 6.1°C to 16°C. This indicates that the flux reflects soil temperature dependence, differing by a factor of nine. The CH4 oxidation rate observed in this study was within the range observed by Whalen et al. .
4.2. Fluxes of CO2, CH4 and N2O in Black Spruce Stand After Forest Fire
 Average fluxes of CO2, CH4 and N2O after the prescribed fire are shown in Table 1. Using data estimated from a black spruce burned in 1996, O'Neill et al.  showed an elevated CO2 emission in burned black spruce after at least three years, which was ascribed to the higher soil temperature (19°C) than that before the burning. This suggests that solar irradiation transfers more easily to the burned stand surface than to the unburned floor containing higher soil-water contents [Kasischke et al., 2000b; Richter et al., 2000]. Richter et al.  reported that the average CO2 flux from black spruce stands burned 10 years ago was 9.5 ± 1.4 mmol/m2/h during the summer season in central Alaska. As the burned forests gradually recover, roots and microbial respiration rates increase progressively with time, leading to the observed higher CO2 flux.
 Most CH4 was oxidized after the fire; however, some CH4 was emitted to the atmosphere, as was observed on 31 August 1999 and 27 September 2000. According to Levin et al., , N2O emission from soils has a positive relationship to fire severity in wetted grass and chaparral ecosystems. The enhanced N2O emission may be ascribed to the result of nitrification induced by enhanced concentrations of soil ammonium [Levin et al., 1988].
 CO2 fluxes showed largely annual variability in mixed hardwood and black spruce stands before the fire, which may be due to low values estimated in early and late summer 1999. On the contrary, fluxes of CH4 and N2O had no annual variability in two boreal forests of central Alaska.
4.3. Effects of Forest Fire on the Fluxes of CO2, CH4 and N2O
 This observation was carried out to uncover the effect of forest fire on the fluxes of CO2, CH4 and N2O before and after the fire. Table 2 shows the effect of the fire on the fluxes of trace gases in black spruce stands.
Table 2. Effect of Forest Fire on Fluxes of CO2, CH4 and N2O Estimated Between Before and After Fires in Black Spruce Stands, Interior Alaska
Soil Temp (°C)
N.D. denotes No Data.
The fluxes cited the values measured in similar black spruce forests before the burning of 1998.
The fluxes used the values obtained before the fire in 1999.
 Forest fires significantly decreased soil CO2 and N2O fluxes by the range of 22 to 50% and the range of 10 to 50%, respectively (Table 2). On the contrary, CH4 flux from the soil increased from 7 to 142%, suggesting that fire plays a role in accelerative thawing of the frozen soil. Subsequently, CH4 may be released from permafrost. The potential post-fire CH4 emission may occur because of the following reasons: First, CH4 emission in 1999 measured at the end of August showed that the heat transfer maximally reached to the permafrost after the fire (K. Yoshikawa, personal communication). Ordinarily, the maximum thawing season for permafrost is the end of September in these black spruce forests. Second, despite low soil temperature (−0.6°C) at the end of September 2000, CH4 flux averaged −19 ± 45 μmol/m2/day. The large standard deviation was due to CH4 emission at three sites and the oxidation at four sites in burned black spruce. Most CH4 after the fire was oxidized to the soil; however, some was emitted to the atmosphere in this burned stand. However, to clarify the CH4 emission after the fire, detailed vertical CH4 profiles and soil profile observations are required in burned black spruce floors.
Kasischke  and Richter et al.  found that the fire decreased the total soil respiration rate by 33 to 59% and 44 to 58%, respectively. The decreased rate in this study was slightly less than their results (22 to 50%). Fires in Alaskan black spruce forests are extremely destructive to all vegetation. Depending on their severity, fires typically consume between 20 and 90% of the organic soil layer, including any living roots present in this layer [Kasischke et al., 2000c].
Schlentner and Van Cleve  estimated that approximately 20% of soil respiration in mature black spruce stands was derived from microbial decomposition and the remainder was from plant root respiration. Using this partition and the total flux of CO2 from chamber observations in a black spruce forest, microbial respiration can be calculated to be 0.8 mmol/m2/h in 1996 and 0.22 mmol/m2/h in 1999 at the unburned sites. These observed respiration rates by flux chambers, which could be solely from microbial respiration in the soil, were 2.9, 0.80 and 0.85 mmol/m2/h in 1996, 1999 and 2000, respectively. These are almost triple the estimated microbial respiration above. This finding indicates that the post-fire condition may stimulate microbial respiration because of higher nutrients and substrates in remnant soil after the burning. Moreover, Van Cleve et al. , Kasischke  and Richter et al.  demonstrated that the forest floor temperatures for boreal forests underlain by permafrost remain warmer than in unburned forests for at least 20 to 30 years after the fire.
4.4. Effect of Soil Temperature on the Fluxes of Trace Gases
 CO2 fluxes significantly depend on soil temperature in two stands (Figures 2a and 2b). In dry weather, in particular, soil temperature plays a crucial role in determining the fluxes in both stands. Under wet conditions, however, there is little particular relation between the soil temperature and the CO2 flux. Q10 is the temperature coefficient of the reaction and is defined as the ratio of reaction rates at an interval of 10°C. Van't Hoff formulated the empirical rule is on the order of 2 to 3. Under fine weather conditions, Q10 values were 3.5 and 3.0 at two black spruce sites, and at sites A and B of mixed hardwood, the values were 4.2 and 2.5, respectively (Table 3). Boone et al.  determined Q10 to be 4.6 for root respiration plus rhizosphere metabolism and 3.5 for respiration by bulk soil at the Harvard Forest. The other studies also have shown Q10 values from 1.3 to 3.3 for soil respiration [Dörr and Münnich, 1987; Raich and Schlesinger, 1992; Lloyd and Taylor, 1994]. As the Q10 value obtained at site A corresponds to that of root respiration reported by Boone et al. , it can be speculated that root respiration is the dominant source of soil CO2 at the site. Whereas, Q10 value at site B indicates that microbial respiration rather than root respiration is predominant.
Table 3. Estimation of Q10 and R Values for the Relationship Between Soil Respiration Rate and Soil Temperature in Mixed Hardwood and Black Spruce Stands During 1998 and 1999, Interior Alaska
 CH4 oxidation rates have an exponential correlation with soil temperature in both stands (Figures 3a and 3b). The pattern of the relationship between CH4 flux and soil temperature is similar to that of CO2 in mixed hardwood stand. CH4 flux has a slight relation to soil temperature in black spruce under dry conditions (r = 0.33), and wet conditions (r = 0.66). CH4 flux from dark chamber observations in black spruce stands showed lower values, which it may be due to the lower temperature within the dark chamber when compared to the light chamber. Born et al.  stated that CH4 flux into soils varied by more than a factor of 10 from one site to another. A previous study on the effect of soil temperature on CH4 flux revealed Q10 values of 1.5 [Born et al., 1990] and 2.0 [Crill, 1991] in temperate soils. The Q10 values in this study varied from 1.4 to 4.1, which were higher than the previously reported values. CH4 flux reflected the significant temperature response only in mixed hardwood sites; however, it did not depend on temperature in black spruce sites.
 N2O flux estimated in this study showed almost exponential increase with temperature (Figures 4a and 4b), and the correlation coefficient (R) ranged from 0.39 to 0.92. The higher N2O fluxes were observed at site A (>2.5 μmol/m2/h, Figure 4a), which is similar to the observed data by Bowden et al. , Schmidt et al. , and Mosier et al.  within the similar temperature range (16°C to 18°C). N2O flux depends on soil temperature and soil moisture; we did not observe the moisture contents in soils. Excluding the higher N2O flux, Q10 values ranged from 1.8 to 2.5 in both stands.
Figure 5 presents the relationship between the fluxes of trace gases and soil temperature at 5 cm below the burned forest floor after the fire. Soil respiration was found to have an exponential correlation to the soil temperature at 5 cm under the surface (Figure 5a). CH4 flux in burned black spruce sites had no relation to the soil temperature; although the flux had a soil-temperature dependency in the burned site of 1996 (r = 0.99, Figure 5b). In spite of the narrow temperature range at the burned 1996 site, CH4 uptake had the soil-temperature dependence, suggesting that microbial activity in burned soil could recover after the fire with time. Van Cleve et al.  indicated the ground layer of a black spruce forest in interior Alaska warmed by 9°C during the growing season for three years, resulting in a 20% loss of ground-layer biomass. The radiation balance of burned black spruce stands changed rapidly following fire and led to a marked increase in soil temperature that might last many years following fire [Richter et al., 2000].
 There are few reports on the relationship between N2O flux and soil temperature after the fire. In this study, the plots indicated an exponential correlation in 1996 (r = 0.99) and 1999 (r = 0.93) observed with both chambers (Figure 5c).
 Relationships between the fluxes of trace gases before and after fire and soil temperature showed good exponential correlations, indicating that soil temperature was one of the factors determining the fluxes of trace gases on boreal forest soils. Therefore, higher soil temperature can be led to reduced soil moisture that can enhance the diffusion of CO2, CH4 and N2O between the atmosphere and terrestrial ecosystem soils, increasing the fluxes of trace gases. So, the effect of temperature may be direct (enhanced microbial activity) and indirect (increased diffusion of trace gases) in black spruce stand after the fire.
4.5. Estimation of Net Respiration Rate by Moss and Lichen Layers
 As described in section 1, the light and dark chambers flux measurements in the forest show the difference between photosynthesis and respiration by mosses and lichens on the black spruce floor. Here, we defined net respiration rate (NRR) by moss and lichen mats as the flux difference estimated by both chambers in black spruce stand floors. Net respiration rates by moss and lichen mats in 1998, 1999, and 2000 before and after the fire are shown in Table 4. In mixed hardwood stands during 1999, the rate is 0.9 mmol/m2/h, which corresponds to 60% of total respiration rate before the fire. In black spruce stands measured before the forest fire, the fraction of respiration rate for total respiration ranged from 42 to 58%. However, after the fire, the net respiration decreased by 20 to 40%, indicating that the rates are not actual respiration by moss and lichen floors, but the respiration by roots and microbes.
Table 4. Estimation of Net Respiration Rate by Moss and Lichen Layers Before and After Burning of 1998, 1999, and 2000 in Black Spruce Forests, Interior Alaska
 The respiration rate of mosses and lichens is found to depend on the soil temperature (R = 0.97, Figure 6), showing the respiration rates significantly depend on the soil temperature at 20 cm depth under the moss and lichen mats in unburned black spruce stands. Moreover, respiration by the layers enables the fluxes estimated by light and dark chambers to be identifiable in Figure 2b regardless of the weather condition. Sveinbjörnsson and Sonesson  found that CO2 concentration levels around shoots of Pleurozium schreberii and Ptillium cristacastrensis in black spruce forest floors in central Alaska were highest in the early morning hours and lowest during midday. Soil respiration and nocturnal moss respiration was the most probable source of the elevated CO2 concentration in black spruce stands.
 The annual average NRR can be estimated to be 55 ± 27 gC/m2/yr on black spruce floors, and 31 ± 14 gC/m2/yr in mixed hardwood based on 120-day snow-free period. Van Cleve and Dryness  reported that an area of interior Alaska dominated by black spruce stands was at most 0.33 × 1012 m2. The regional net respiration rate by moss and lichen mats on black spruce floors of central Alaska was maximum 0.018 ± 0.009 GtC/yr, indicating that the respiration rate of moss-lichen mats was one of the sources of atmospheric CO2 in boreal forests.
4.6. Implications for the Regional Carbon Budget After the Fire
 Based on the proportionality of Schlentner and Van Cleve , the total soil respiration rate in 1-year-old burned black spruce is estimated to be 0.85 mmol/m2/h. Of this, 0.63 mmol/m2/h are attributed to stimulation of microbial decomposition after the fire. Over a 120-day growing period, this stimulation amounts to a soil carbon flux of 0.22 tC/ha. Including the additional flux in the months between summer and winter and presuming half the rate of the growing season flux (0.32 mmol/m2/h for 90-day transition period), the microbial respiration might be stimulated by as much as 0.30 tC/ha in a black spruce stand after a forest fire. Because permafrost melts and mineral soil profiles increase for over a decade following fire, microbial respiration rate may be on the order of 3.0 tC/ha from the terrestrial ecosystem to the atmosphere. On the other hand, the total soil respiration in black spruce burned in 1996 is evaluated to be 2.9 mmol/m2/h. 2.1 mmol/m2/h are ascribed to prompting of microbial respiration decomposition after forest fire. Over three years after the fire, the microbial respiration was 3.5 tC/ha. The rate can be estimated 11.7 tC/ha over a decade in post-burn black spruce stands. The difference of microbial respiration in two sites results from a magnitude of fire severity that is closely related to successional chronosequence, soil temperature and soil moisture [Van Cleve et al., 1983; Kasicshke, 2000; Kasischke et al., 2000c]. Burned black spruce stands play a crucial role as a source of atmospheric CO2 after a fire (post-fire microbial respiration of 22.2 tC/ha [Richter et al., 2000]). Fire severity is linked with the long-term patterns of cooling and warming that have occurred in this region. Therefore, long-term patterns of post-fire ground temperature and moisture in boreal forests are not in a steady state, but vary in response to change in climate and fire severity.
 We are grateful to Professor M. Fukuda for help with the field observation during FROSTFIRE burning experiment. We would like to express our gratitude to the staff and all the students in the laboratory of Marine Atmospheric Geochemistry (MAG), Hokkaido University. Editor (Dr. D. Stillman) and two anonymous reviewers are thanked for helpful comments on this manuscript. Also, we appreciate the financial support from the Frontier Observational Research System for Global Change/International Arctic Research Center to the author (Yongwon KIM) during his term to complete the manuscript. A part of this study (2000 observation) was supported by National Space Development Agency of Japan (NASDA) through NASDA Arctic Research Projects using IARC-NASDA Information System (INIS).