Variation of Hydrological Regime with Permafrost Coverage Over Lena Basin in Siberia Baisheng Ye,1 Daqing Yang,2 Zhongliang Zhang,3 and Douglas L

Home , Aldan (river), Baikal Mountains, Yana (river)

JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 114, D07102, doi:10.1029/2008JD010537, 2009 Click Here for Full Article

Variation of hydrological regime with permafrost coverage over Lena Basin in Siberia Baisheng Ye,1 Daqing Yang,2 Zhongliang Zhang,3 and Douglas L. Kane2 Received 3 June 2008; revised 18 November 2008; accepted 4 February 2009; published 2 April 2009.

[1] We use monthly discharge and permafrost data to examine the relationship between discharge characteristics and basin permafrost coverage for the nested subbasins of the Lena River in Siberia. There are similarity and variation in streamflow regimes over the basin. The ratios of monthly maximum/minimum flows directly reflect discharge regimes. The ratios increase with drainage area from the headwaters to downstream within the Lena basin. This pattern is different from the nonpermafrost watersheds, and it clearly reflects permafrost effect on regional hydrological regime. There is a significant positive relationship between the ratio and basin permafrost coverage. This relationship indicates that permafrost condition does not significantly affect streamflow regime over the low permafrost (less than 40%) regions, and it strongly affects discharge regime for regions with high permafrost (greater than 60%). Temperature and precipitation have similar patterns among the subbasins. Basin precipitation has little association with permafrost conditions and an indirect relation with river flow regimes. There exists a good relation between the freezing index and permafrost extent over the basin, indicating that cold climate leads to high coverage of permafrost. This relation relates basin thermal condition with permafrost distribution. The combination of the relations between temperature versus permafrost extent, and permafrost extent versus flow ratio links temperature, permafrost, and flow regime over the Lena basin. Over the Aldan subbasin, the maximum/minimum discharge ratios significantly decrease during 1942–1998 due to increase in base flow; this change is consistent in general with permafrost degradation over eastern Siberia. Citation: Ye, B., D. Yang, Z. Zhang, and D. L. Kane (2009), Variation of hydrological regime with permafrost coverage over Lena Basin in Siberia, J. Geophys. Res., 114, D07102, doi:10.1029/2008JD010537.

1. Introduction permafrost changes, and human impacts [Peterson et al., 2002; Yang et al., 2002; Ye et al., 2003; McClelland et [2] River runoff is the primary freshwater source to the al., 2004]. Arctic Ocean. Fresh water discharge from the northern- [3] In the cold regions, hydrological regime is closely flowing rivers plays an important role in regulating the related with permafrost conditions, such as permafrost thermohaline circulation of the world’s oceans [Aagaard extent and thermal characteristics. Ice-rich permafrost has and Carmack, 1989]. Both the amount and the timing of a very low hydraulic conductivity and commonly acts as a freshwater inflow to the ocean systems are important to barrier to deeper groundwater recharge or as a confining ocean circulation, salinity, and sea ice dynamics [Aagaard layer to deeper aquifers. Because it is a barrier to recharge, and Carmack, 1989; Macdonald, 2000]. Recent studies permafrost increases the surface runoff and decreases sub- report significant changes in arctic hydrologic system, surface flow. Permafrost extent over a region plays a key particularly cold season and annual discharge increases over role in the distribution of surface-subsurface interaction large Siberian watersheds. These changes indicate hydro- [Lemieux et al., 2008; Carey and Woo, 2001; Woo et al., logic regime shifts due to large-scale climate variations, 2008]. Permafrost and nonpermafrost rivers have very different hydrologic regimes. Relative to nonpermafrost basins, permafrost watersheds have higher peak flow and 1States Key Laboratory of Cryospheric Sciences, Cold and Arid Regions Environmental and Engineering Research Institute, Chinese Academy of lower base flow [Woo, 1986; Kane, 1997]. In the permafrost Sciences, Lanzhou, China. regions, watersheds with higher permafrost coverage have 2Water and Environment Research Center, Civil and Environmental lower subsurface storage capacity and thus a lower winter Engineering Department, University of Alaska Fairbanks, Fairbanks, base flow and a higher summer peak flow [Woo, 1986; Alaska, USA. 3Department of Resources and Environment, Lanzhou University, Kane, 1997; Yang et al., 2003]. Lanzhou, China. [4] It is difficult to accurately determine changes in permafrost conditions. Our understanding of permafrost Copyright 2009 by the American Geophysical Union. change and its effect on hydrological regime is incomplete. 0148-0227/09/2008JD010537$09.00

D07102 1of12 D07102 YE ET AL.: STREAMFLOW REGIME VERSUS PERMAFROST COVERAGE D07102

Figure 1. The Lena River watershed and hydrological stations (A through K) used in this study. Also shown are the reservoir location on the Vilui River and the permafrost distribution [Brown et al., 1997, 2001] over the Lena basin.

For instance, there are uncertainties regarding the impact of 15% of the total freshwater flow into the Arctic Ocean ground ice melt and its contribution to annual flow changes [Yang et al., 2002; Ye et al., 2003]. Relative to other large over large Siberian rivers [McClelland et al., 2004; Zhang et rivers, the Lena basin has less human activities and much al., 2005a]. Permafrost condition and streamflow character- less economic development [Dynesius and Nilsson, 1994]. istics vary within large watersheds in Siberia. Examination There is only one large reservoir in the Vilui subbasin. A and comparison of hydrological regimes between subbasins large dam (storage capacity 35.9 km3) and a power plant with various permafrost conditions can improve our under- were completed in 1967 near the Chernyshevskyi standing of impact of permafrost changes on cold region (112°150W, 62°450N). This reservoir is used primarily for hydrology. This paper examines the relationship between electric power generation: holding water in spring and hydrological regime and permafrost coverage over nested summer seasons to reduce snowmelt and rainfall floods subbasins within the Lena River in Siberia. It analyzes and releasing water to meet the higher demand for power in monthly discharge data, with a focus on the ratio of the winter [Ye et al., 2003]. Various type of permafrost exists maximum to minimum discharge (Qmax/Qmin) and its in the Lena basin, including sporadic, or isolated permafrost relation with permafrost condition, because this ratio in the source regions, and discontinuous and continuous reflects the hydrological regime. The objective of this study permafrost in downstream regions (Figure 1) [Brown et al., is to quantify the impact of permafrost on streamflow 1997]. Approximately 78–93% of the Lena basin is under- regime and change, and to specifically define a relationship lain by permafrost [Zhang et al., 1999; McClelland et al., between basin permafrost extent and streamflow conditions 2004]. over the Lena watershed. We also examine relationship [6] Since the late 1930s hydrological observations in the between basin air temperature and precipitation and their Siberian regions, such as discharge, stream water tempera- effect on permafrost extent and basin streamflow regimes. ture, river ice thickness, dates of river freezeup and breakup, The result of this study will shed light on our knowledge of have been carried out systematically by the Russian Hydro- cold region hydrology and its change due to climate impact meteorological Services and the observational records and human influence. were quality-controlled and archived by the same agency [Shiklomanov et al., 2000]. The discharge data are now 2. Basin Description, Data Sets, and Method available from the R-ArcticNet (v4.0), a database of Pan- of Analysis Arctic River Discharge during 1936–2000 [Lammers et al., 2001]. In this analysis, long-term monthly discharge records [5] The Lena River originates from the Baikal Mountains collected at 9 stations (A–H and K) along the main stem in the south central Siberian Plateau and flows northeast and and 2 tributary stations (I and J) were used (Figure 1). north, entering into the Arctic Ocean via the Laptev Sea Relevant information of these stations is given in Table 1. 2 (Figure 1). Its drainage area is about 2,430,000 km , mainly [7] Permafrost data and information were obtained from covered by forest and underlain by permafrost. The Lena the database of the digital permafrost map compiled by River contributes 524 km3 of freshwater per year, or about International Permafrost Association (IPA) [Brown et al.,

2of12 D07102 YE ET AL.: STREAMFLOW REGIME VERSUS PERMAFROST COVERAGE D07102

Table 1. List of Hydrologic Stations Used in This Studya

Station Code Latitude Longitude Drainage Area Annual Runoff (see Figure 1) Station Name/Location (°N) (°E) Data Period (1000 km2) %km3 mm % A Kachug 53.97 105.88 1936 1990 17 0.7 2.9 164.0 0.5 B Zhigalovo/Upper Lena 54.82 105.13 1969 1977 30 1.3 3.8 124.6 0.7 C Gruznovka/Upper Lena 55.13 105.23 1912 1990 42 1.7 6.2 148.0 1.2 D Ust’-Kut/Upper Lena 56.77 105.65 1936 1986 71 2.9 10.1 141.9 1.9 E Zmeinovo/Upper Lena 57.78 108.32 1936 1990 140 5.8 35.4 253.1 6.7 F Krestovskoe/Upper Lena 59.73 113.17 1936 1999 440 18.1 131.3 298.5 24.8 G Solyanka/Upper Lena 60.48 120.7 1933 1999 770 31.7 213.5 277.2 40.4 H Tabaga/Upper Lena 61.83 129.6 1936 1999 897 36.9 221.0 246.4 41.8 I Verkhoyanskiy Perevoz/ 63.32 132.02 1942 1999 696 28.6 166.0 238.5 31.4 Aldan subbasin outlet J Vilyuy/Vilui valley outlet 63.95 124.83 1936 1998 452 18.6 46.7 103.3 8.8 K Lena At Kusur 70.68 127.39 1934 2000 2430 100.0 528.6 217.5 100.0 aStations A–H are listed from upstream to downstream.

1997, 2001]. The shapefiles were derived from the original itation and temperature and their relationships with basin 1:10,000,000 printed map. The IPA permafrost map was permafrost extent and streamflow regime. categorized as continuous (90–100%), discontinuous (50– 90%), isolated (10–50%), and sporadic (0–10%). The 3. Results coverage of permafrost (CP) in a basin has been determined 3.1. Hydrographs differently [McClelland et al., 2004; Zhang et al., 1999]. Most studies consider permafrost existence [Smith et al., [9] Figure 2 shows the mean monthly discharge at the 2007; McClelland et al., 2004] and determined the total 9 stations from the upper to lower Lena basin. They coverage of permafrost (4 types of permafrost) [McClelland generally have similar runoff regimes, i.e., high discharge et al., 2004]. In this study, basin and subbasin permafrost in summer and low flow in winter. However, the difference areas are determined by overlaying the permafrost map to among them is very obvious. The large difference is seen in the river network map. Basin/subbasin boundaries and the peak flow, both in rate and timing. Lena river peak drainage areas are derived from 1 km DEM data, the USGS discharge is mainly related to snowmelt [Yang et al., 2007; Ye et al., 2003], as it varies over the basin. The mean peak GTOPO30 data set [Bliss and Olsen, 1996]. The drainage 3 areas from the 1km DEM match well to those reported by flow is about 210 m /s at the Kachug station located in the upper basin (station A in Figure 1), due to increasing drainage the Pan-Arctic River Discharge Database [Lammers et al., 3 2001], with the relative errors being less than 15% for the area, it gradually grows to 74,000 m /s at the basin outlet subbasins. The coverage of permafrost in a basin is defined (the Kusur station, K in Figure 1). The monthly peak flow as the weighed average of the 4 different types of perma- occurs in May in the upstream source regions (stations A–D) frost. Considering the ranges of permafrost coverage, we and in June over the downstream regions (stations E–H take the mean coverage as representative coverage for every and K), because early snowmelt in the source regions and permafrost type, i.e., 95%, 70%, 30% and 5% permafrost later melt downstream [Yang et al., 2003, 2007]. coverage in continuous, discontinuous, isolated and sporadic [10] As expected, the minimum discharge increased with areas, respectively. Variations from the mean CP are ±5%, drainage area in the Lena basin. The lowest monthly flow is about 22 m3/s at the Kachug station (A in Figure 1) and it ±20%, ±20% and ±5% for continuous areas, discontinuous 3 areas, isolated areas, and sporadic permafrost, respectively. gradually increases to 1100 m /s at the basin outlet (K in [8] The methods of analyses include calculation of Figure 1). The lowest flow occurs in February in the monthly mean discharge and hydrographs for the nested headwaters (stations A–E), in March over the middle subbasins within the Lena watershed, determinations of streams (F–G), and in April in downstream (stations H ratio of monthly high to low flows, and examination of and K). This pattern, early low flow in the source regions the relation between the ratio and permafrost coverage over and late low flow in lower basin, is related with the timing the nested subbasins. We compare the ratio between pre-dam of snowmelt and hydraulic routing of flow. Also, the timing and post-dam periods to quantify and discuss reservoir of snowmelt usually dictates the timing of discharge reces- impact on streamflow regime and change. We conduct trend sion. Usually early (late) snowmelt is associated with early analysis to define the long-term changes in hydrographs and (late) peak and subsequently early (late) low flow. Figure 2 use the Student t test to determine the statistical significance clearly illustrates the variation of hydrographs over the Lena of the trends. Furthermore, we compare the ratio trends basin. The difference in hydrographs reflects basin climatic among the subbasins to understand the regional differences and physical characteristics, particularly permafrost condi- in streamflow characteristics and changes. We also use tions. Among other measures, the high and low flows monthly precipitation (P) and temperature (T) data [Jones, determine the hydrograph shape. The ratio of monthly 1994] for the Siberian regions and calculate the basin mean maximum discharge (Qmax) to minimum discharge (Qmin) values for the upper Lena, Aldan and Vilyuy subbasins. is a direct measure of hydrograph shape and hydrologic Based on these data, we carry out analyses of basin precip- regime [Woo, 1986; Kane, 1997]. This ratio can also indicate streamflow regime changes caused by reservoir

3of12 D07102 YE ET AL.: STREAMFLOW REGIME VERSUS PERMAFROST COVERAGE D07102

Figure 2. Monthly hydrographs at selected stations along the main river valley. regulation [Ye et al., 2003; Yang et al., 2004a, 2004b]. Ice basin (Figure 1). The extent of permafrost gradually jams in the northern regions affect hydrograph in spring. In increases from 20% in the headwaters to 86% in down- this study, we use monthly maximum and minimum dis- stream regions (Figure 3). The Qmax/Qmin ratios are low charge to determine both the hydrograph shape and the (10–20) for CP less than 40%, and moderate (20–30) for calculation of the ratio. We did not use any daily (maximum CP interval of 60–75%, and high (about 70–80) for CP and minimum) discharge. This may help to eliminate or greater than 86%. The ratios change with CP over the basin, smooth the ice jam effect on flow regime analysis. i.e., increasing with the CP from upstream to downstream. [11] Calculations of the Qmax/Qmin for the 9 stations in [13] Dam regulation can significantly alter streamflow the Lena basis show variations from upstream to down- regime [Ye et al., 2003; Yang et al., 2004a, 2004b]. It is stream. The mean Qmax/Qmin are about 13–16 in the source therefore critical to use pre-dam and natural flow data for regions (stations A–E), and increases from 21 in middle regional hydrology analyses. Using the flow data from the upper Lena (station F) to 74 at the outlet station (stations K). upper Lena without dam regulation and the pre-dam data for The ratio increases with drainage area (Figure 3). This the basin outlet, we conducted regression analysis between positive relationship is different from nonpermafrost the Qmax/Qmin and CP. The result reveals a significant regions/basins, where the ratio decreases with basin area. relationship (significant at 99%, Figure 4). This relationship clearly indicates the effect of permafrost distribution on 3.2. The Relationship Between Qmax/Qmin Ratio and discharge process. Permafrost regions/basins have low in- Coverage of Permafrost (CP) filtration and limited subsurface storage capacity [Woo, [12] The permafrost distribution varies from sporadic in 1986; Kane, 1997; Woo et al., 2008]. Relative to non- the source regions to continuous permafrost over the Lena permafrost or low permafrost basins, high permafrost basins

Figure 3. The mean ratio of Qmax/Qmin and coverage of permafrost (CP) versus drainage areas. The ratio for Kusur (station K) is for pre-dam data.

4of12 D07102 YE ET AL.: STREAMFLOW REGIME VERSUS PERMAFROST COVERAGE D07102

Figure 4. The ratios of Qmax/Qmin versus coverage of permafrost (CP) at the nine stations from the upper Lena to basin outlet without reservoir regulation. Two additional stations (the Aldan and Vilui subbasins in Lena basin) and one additional basin (the Yasachnaya Kolyma River) in East Siberian are also shown. The ratios of Qmax/Qmin are showed at the Vilui subbasin and Kusur at Lena basin outlet with reservoir regulation since 1967. Note: the Y axis is in logarithm scale. have higher maximum discharge and lower minimum dis- dam period. The higher ratio for the Vilui is due to charge, hence higher ratio of Qmax/Qmin. Figure 4 shows continuous permafrost and very low winter flows in the that the permafrost condition does not significantly affect northern parts of the Lena basin. The ratios and CP of these streamflow regime over the low permafrost regions (i.e., CP two subbasins are generally consistent with the relationship, < 40%), and it strongly affects discharge regime for regions showing higher Qmax/Qmin ratios with higher CP. Further- with high permafrost coverage (i.e., CP greater than 60%). more, we include the Yasachnaya Kolyma watershed in It is useful to determine this threshold of permafrost impact eastern Siberia for a comparison. This basin has high to regional hydrology, particularly in regions undergoing permafrost extent (about 94%) and fits well with the curve climate warming and permafrost degradation. Further anal- derived from the Lena basin (Figure 4). These results may yses over broader regions are necessary to refine and verify indicate a general applicability of the relation over the the relation and threshold. northern basins. It is, however, important to recognize the [14] The Qmax/Qmin–CP relation has been derived for limitation of this relation. The Qmax/Qmin ratios can be the nested basins from the upper Lena to the basin as a very high due to very low base flow over continuous whole. To test the relationship, two independent subbasins permafrost regions. For instance, winter base flow is ex- within the Lena watershed (i.e., the Aldan and Vilui sub- tremely low in the very cold Arctic coast watersheds with basins, stations I and J in Figure 1) have been included in almost 100% permafrost coverage, such as the Yana and Figure 4. These two tributaries have very high CP, 92% Olenka basins near the Lena River. and 95% for the Aldan and Vilui, respectively. There is a reservoir in the upper Viliu valley, with a storage capacity of 3.3. Qmax/Qmin Ratio Changes Over Time 36 km3, or about 77% and 7% of annual discharge at Vilui [16] It is difficult to survey permafrost change over large outlet and Lena basin mouth. Operation of this dam has regions. It is easy to calculate the ratio of Qmax/Qmin over changed streamflow regimes [Ye et al., 2003]; it reduced the a basin where discharge data is available. The relationship peak flows by about 30% and enhanced low flows by about between the Qmax/Qmin ratio and CP directly links stream- 27 times at Vilui outlet (station J in Figure 1). Because of flow regime with basin permafrost coverage. The changes in these changes, the ratio of Qmax/Qmin sharply decreases the ratio over time and space perhaps can, therefore, reflect from 536 to 38 at the Vilui outlet (station J in Figure 1) and basin permafrost changes. from 75 to 45 at the Lena outlet (K in Figure 1), respectively [17] Figure 5 shows the time series of Qmax/Qmin ratio (Figure 4). during 1935–2000 at 3 tributary stations and at the Lena [15] The Qmax/Qmin ratios are high, 75 for the Aldan river outlet. The ratio at the Tabaga station (H) in the upper (Station I) and 536 for the Vilui (Station J) during the pre- Lena varies from 16 to 48, with a mean of 28. The ratio does

5of12 D07102 YE ET AL.: STREAMFLOW REGIME VERSUS PERMAFROST COVERAGE D07102

Figure 5. The ratio of Qmax/Qmin (a) at stations of Tabaga in upper Lena and Verkhoyanskiy Perevoz in Aldan without dam regulation and (b) at Viluy in Vilui and Kusur at Lena outlet during 1935–2000. not show any trend, although it is obviously lower during over the study period and the total trend was 114% for the 1980–1998, with a mean of 26 due to higher Qmin. The Qmin during 1942–1999 (Figures 6a and 6b). Low flow ratio in the Aldan subbasin ranges from 24 to 177, with a changes are particularly strong during 1976–1999, with a mean of 75. It has significantly (95% confidence) decreased step rise from 200–400 m3/s to 600–800 m3/s in April. by 54 (75%) during 1942–1998 due to a strong increase There are large fluctuations in the low flow records over the in Qmin. There are no dams in this tributary. The decrease study period. It is important to note the consistency in the of the ratio is mainly due to increases in winter flows [Ye et interannual variation of low flow among the months, i.e., al., 2003]. The ratio for the Vilui tributary varies between 6 high flows in early winter translating to high flows in late and 1,000. It has sharply decreased since 1966 when the winter. dam went operational. The mean ratios are 531 and 15 for [19] In addition, regression analyses of the low flows pre- and post-1967, respectively. The drop in ratio is caused show high correlations (R2 = 0.73 to 0.94) between the by the operation and regulation of the Vilui reservoir around consecutive months during November to April (Figure 7). 1967 [Ye et al., 2003], i.e., enhancing winter flows and This result is expected, as the winter hydrology is domi- reducing summer flows downstream of the dam. The ratio at nated by discharge recession. the Lena river outlet varies from 24 to 164, with a mean of [20] The ratio of monthly flows between consecutive 60. It decreases over time, with the total trend of 58 months, or monthly recession coefficient (RC), is a quan- (nearly 95%) at 99% confidence during 1935–2000. This titative measure of recession processes. The mean monthly decrease is mainly caused by reservoir regulation in the recession coefficients for this subbasin ranges from 0.62 in Vilui valley. early winter (December/Nov) to 0.92 in later winter (Apr/ [18] Over the Lena basin, the Qmax/Qmin ratio has a very March), indicating fast (slow) recession in early (late) strong decreasing trend for the Aldan valley. This change in winter (Figure 7). The time series of recession coefficient ratios depends on the max and min discharge. Monthly flow over the study period shows significant increases (at 99% data (Figures 6a and 6b) show that peak flow did not change confidence) in most winter months, except an insignificant much, and the low flows during November to April in- decrease for December/November (Figure 8). Increasing creased significantly by 50–117%, (99% significant level) RC over time suggests decreases in recession rate. Dis-

6of12 D07102 YE ET AL.: STREAMFLOW REGIME VERSUS PERMAFROST COVERAGE D07102

Figure 6. (a) The maximum monthly discharge and (b) the monthly discharge during low-flow months during cold season at Aldan station during 1942–1999. charge recession is generally a process of storage release. the 3 subbasins within the Lena watershed (Figure 10a), Studies report significant permafrost changes in this region, although the discharge patterns differ among the subbasins. including warming of ground temperatures and deepening This association of precipitation with flow regime is rea- of active layer depth [Pavlov, 1994; Frauenfeld et al., 2004; sonable, as in the northern regions and watersheds, snowfall Zhang et al., 2005b; Romanovsky et al., 2007]. There seems accumulates over winter season and snowpack melt in to be a general coherency between hydrology and perma- spring produces the highest floods. frost changes over this region, which deserves more re- [23] Monthly mean temperatures also have similar pat- search attention. terns among the subbasins (Figure 10b). To relate temperature with permafrost conditions, previous studies use 4. Discussion various temperature indexes, including seasonal and yearly means and the freezing index (FI). FI is defined as the [21] In addition to permafrost conditions, other factors, cumulative degree days of negative temperature over a year. such as geography, geology, and climate also influence This index is often used to calculate the freezing depth streamflow regimes in the northern regions. Data availabil- [Nelson and Outcalt, 1983, 1987]. Calculations of the FI, ity is a limitation for most arctic research, including this using basin mean monthly temperature data, show a range study. To understand the climate effect on permafrost and from 140 to 175°C for the 3 tributaries. basin hydrology, we have obtained monthly precipitation [24] There is a good relation between the FI and perma- and temperature data from CRU [Jones, 1994] for the frost extent (coverage of permafrost, CP), i.e., CP increase Siberian region and calculated the basin mean values for with FI for the 3 tributaries (Figure 11a). This result, as the upper Lena, Aldan, and Vilyuy subbasins. Using these expected, indicates cold climate leads to high coverage of data, we examine precipitation and temperature patterns permafrost. This relation is very important, as it relates over the subbasins and their relationships with basin per- basin thermal condition with permafrost distribution. Our mafrost extent and streamflow regime. analyses have already established the relationship between [22] Results of analyses of basin precipitation data show permafrost extent and flow regime. The combination of little association of precipitation with permafrost conditions relations between T-CP and CP ratio links temperature, (Figure 9), this is because temperature mostly controls permafrost, and flow regime over the Lena basin. large-scale permafrost distribution [Nelson and Outcalt, [25] In addition, the Qmax/Qmin ratio increases with the 1983]. We discovered similar precipitation patterns over FI (Figure 11b) over the Lena basin. This is reasonable, as

7of12 D07102 YE ET AL.: STREAMFLOW REGIME VERSUS PERMAFROST COVERAGE D07102

[26] This study found significant winter flow increase over the Aldan subbasin. Other analyses also reported base flow rise over the large northern watersheds, such as the Lena basin [Ye et al., 2003] and Yukon river [Walvoord and Striegl, 2007]. Base flow changes are likely related with changes in subsurface processes and frozen ground conditions in northern regions. Liu et al. [2003] documented significant of winter discharge in the source areas of the Songhuajiang river with high permafrost extent in the northeast China, and related this change to the warming of ground temperatures and deepening of active layer depth. Studies show that active layer increase affects runoff characteristics in a small headwater catchment (Mogot River) in the southern mountainous region of eastern Siberia [Yamazaki et al., 2006] and in the Kuparuk River basin [McNamara et al., 1998]. McNamara et al. [1998] reported a long recession to the slow drainage of water from a thawing active layer in northern Alaska. These changes in base features may be due to changes in storage capacity and amount in the thawed active layer [Woo et al., 2008]. [27] There is little direct measurement of basin storage capacity and amount over the northern regions. It is therefore difficult to accurately determine their changes over time. The basin storage capacity is a measure of groundwater reservoir. A large storage capacity leads to a slow recession. For a given storage capacity, a large storage amount results in a fast recession. Over the Aldan subbasin, the recession slows down from early winter to early spring (Figure 7) maybe due to decrease in storage amount over winter. River discharge increased over the Aldan basin in winter during 1940–1999 (Figure 6b), suggesting storage amount did increase over time. This increase should lead to a fast recession, but, in fact, recession slows down in most winter months (Figure 8). This may imply the increase of basin storage capacity. In addition, Ye et al. [2003] found that discharge increase in November is the strongest over the winter months over the Lena basin. This suggests more storage of water in November over the basin and a fast recession. However, RC decreased in December/November, this may imply that the negative effect of increased storage amount on RC may overcome positive effect of increased storage capacity. More research is necessary to investigate the linkage between changes in subsurface storage and river streamflow.

Figure 7. The relationship of discharge between the 5. Conclusions consecutive months in low-flow periods at Aldan during [28] Monthly discharge regimes vary over the Lena basin, 1942–1999. although it is generally high in summer and low in winter. The ratios of Qmax/Qmin are a direct measure of hydrologic regime. They increase with drainage area from the rivers in cold regions have very low base flows and very headwaters to downstream within the Lena basin. This high peak flows, consequently high Qmax/Qmin ratio. It is pattern is different from the nonpermafrost watersheds. important to mention that monthly T data have been used in The ratios also change with basin permafrost coverage, this analysis for the determination of FI over the subbasins. suggesting that permafrost affects regional hydrology. Re- There might be uncertainties in the FI calculations particu- gression analysis between the Qmax/Qmin ratio and basin larly during the early and late winter periods when the permafrost coverage reveals a significant relationship at monthly T is close to 0C. Daily T data can better define the 99% confidence. This relationship quantifies the effect of FI, but the data was not available to this study for the Lena permafrost distribution on discharge process. It shows that basin. It is expected that the relationship of T with CP and permafrost conditions do not significantly affect streamflow Qmax/Qmin ratio may improve when daily T data are used regime over the low permafrost (less than 40%) regions, and in future analyses. it strongly affects discharge regime for regions with high

8of12 D07102 YE ET AL.: STREAMFLOW REGIME VERSUS PERMAFROST COVERAGE D07102

Figure 8. The monthly recession coefficient between the consecutive months in low-flow period at Aldan during 1942–1999.

9of12 D07102 YE ET AL.: STREAMFLOW REGIME VERSUS PERMAFROST COVERAGE D07102

(greater than 60%) permafrost coverage. It is important to define this relation of permafrost impact to regional hydrology, as this knowledge will be useful for further investiga- tion into permafrost effect on basin hydrology over the large northern regions. Temperature and precipitation have similar patterns among the subbasins. Basin precipitation has little association with permafrost conditions. There is a reasonable relation between the FI and permafrost extent over the basin, indicating that cold climate leads to high coverage of permafrost. This relation relates basin thermal condition with permafrost distribution. The combination of the relations between T-CP and CP-Qmax/Qmin links temperature, permafrost, and flow regime over the Lena basin. [29] The variations in the Qmax/Qmin ratios reflect hydrologic regime changes. Trend analyses of the ratios show different results over the Lena basin. The ratios significantly decrease in the Vilui valley and at the Lena Figure 9. The basin annual precipitation versus CP for the basin outlet due to dam regulation. In the unregulated upper three tributaries (Upper Lena, Aldan and Vilui) in the Lena Lena regions, the ratios do not show any trend during River basin.

Figure 10. (a) The basin monthly mean precipitation and (b) the temperature for three tributaries in Lena River.

10 of 12 D07102 YE ET AL.: STREAMFLOW REGIME VERSUS PERMAFROST COVERAGE D07102

Figure 11. The freezing index (FI) for October to April versus (a) CP and (b) Qmax/Qmin ratio for three tributaries (Upper Lena, Aldan and Vilui) in Lena River basin.

1936–1998. This may imply that the permafrost changes in Bliss, N. B., and L. M. Olsen (1996), Development of a 30-arc-second digital elevation model of South America, in Pecora Thirteen, Human the region were not sufficient enough to affect hydrological Interactions With the Environment—Perspectives From Space,Sioux processes. In the Aldan tributary without regulation, the Falls, South Dakota, August 20–22. ratios significantly decrease due to increase in base flow Brown, J., O. J. Ferrians Jr., J. A. Heginbottom, and E. S. Melnikov (1997), Circum-Arctic map of permafrost and ground-ice conditions, in Circum- during 1942–1998. This change is consistent with perma- Pacific Map Series CP-45, scale 1:10,000,000, 1 sheet, U.S. Geol. Surv. frost degradation over the eastern Siberia. This consistency in Cooperation with the Circum-Pacific Council for Energy and Mineral to some extent verifies permafrost influence on basin Resources, Washington, D. C. streamflow process and change. Our efforts continue to Brown, J., O. J. Ferrians Jr., J. A. Heginbottom, and E. S. Melnikov (2001), Circum-Arctic map of permafrost and ground-ice conditions, Digital refine this relationship for various climate and permafrost Media, Natl. Snow and Ice Data Cent., Boulder, Colo. conditions in the high latitude regions. Carey, K., and M. Woo (2001), Slope runoff processes and flow generation in a subarctic, subalpine catchment, J. Hydrol., 253, 110–129. Dynesius, M., and C. Nilsson (1994), Fragmentation and flow regulation [30] Acknowledgments. This study was supported by National Basic of river systems in the northern third of the world, Science, 266, 753– Research Program of China (2007CB411502), Innovation Project of CAS 762. (KZCX3-SW-345), USA NSF grants 0230083, 0612334 and 0335941 and Frauenfeld, O. W., T. Zhang, R. G. Barry, and D. Gilichinsky (1994), International Partnership Project of CAS (CXTD-Z2005-2). Interdecadal changes in seasonal freeze and thaw depths in Russia, J. Geophys. Res., 109, D05101, doi:10.1029/2003JD004245. References Jones, P. D. (1994), Hemispheric surface air temperature variations: A Aagaard, K., and E. C. Carmack (1989), The role of sea ice and other fresh reanalysis and an update to 1993, J. Clim., 7, 1794–1802. water in the arctic circulation, J. Geophys. Res., 94(C10), 14,485– Kane,D.L.(1997),TheimpactofArctichydrologicperturbationson 14,498. Arctic ecosystems induced by climate change, in Global Change and

11 of 12 D07102 YE ET AL.: STREAMFLOW REGIME VERSUS PERMAFROST COVERAGE D07102

Arctic Terrestrial Ecosystems, Ecol. Stud., vol. 124, edited by W. C. Woo, K., D. Kane, S. Carey, and D. Yang (2008), Progress in Permafrost Oechel et al., pp. 63–81, Springer, New York. Hydrology in the New Millennium, Permafr. Periglac. Process., 19, Lammers, R., A. Shiklomanov, C. Vorosmarty, B. Fekete, and B. Peterson 237–254. (2001), Assessment of contemporary arctic river runoff based on obser- Yamazaki, Y., J. Kubota, T. Ohata, V. Vuglinsky, and T. Mizuyama (2006), vational discharge records, J. Geophys. Res, 106(D4), 3321–3334. Seasonal changes in runoff characteristics on a permafrost watershed in Lemieux, J.-M., E. A. Sudicky, W. R. Peltier, and L. Tarasov (2008), the southern mountainous region of eastern Siberia, Hydrol. Processes, Dynamics of groundwater recharge and seepage over the Canadian land- 20, 453–467. scape during the Wisconsinian glaciation, J. Geophys. Res., 113, F01011, Yang, D., D. Kane, L. Hinzman, X. Zhang, T. Zhang, and H. Ye (2002), doi:10.1029/2007JF000838. Siberian Lena river hydrologic regime and recent change, J. Geophys. Liu, J., N. Hayakawab, M. Lub, S. Dongc, and J. Yuan (2003), Hydrolo- Res., 107(D23), 4694, doi:10.1029/2002JD002542. gical and geocryological response of winter streamflow to climate warm- Yang, D., D. Robinson, Y. Zhao, T. Estilow, and B. Ye (2003), Streamflow ing in Northeast China, Cold Reg. Sci. Technol., 37, 15–24. response to seasonal snow cover extent changes in large Siberian water- Macdonald, R. W. (2000), Arctic estuaries and ice: A positive-negative sheds, J. Geophys. Res., 108(D18), 4578, doi:10.1029/2002JD003149. estuarine couple, in The Freshwater Budget of the Arctic Ocean: Pro- Yang, D., B. Ye, and A. Shiklomanov (2004a), Streamflow characteristics ceedings of the NATO Advanced Research Workshop, edited by E. L. and changes over the Ob river watershed in Siberia, J. Hydrometeorol., Lewis et al., pp. 383–407, Springer, New York. 5(4), 69–84. McClelland, J. W., R. M. Holmes, B. J. Peterson, and M. Stieglitz (2004), Yang, D., B. Ye, and D. Kane (2004b), Streamflow changes over Siberian Increasing river discharge in the Eurasian Arctic: Consideration of Yenisei river basin, J. Hydrol., 296(1–4), 59–80. dams, permafrost thaw, and fires as potential agents of change, J. Geophys. Yang, D., Y. Zhao, R. Armstrong, D. Robinson, and M.-J. Brodzik (2007), Res., 109, D18102, doi:10.1029/2004JD004583. Streamflow response to seasonal snow cover mass changes over large McNamara, J. P., D. L. Kane, and L. D. Hinzman (1998), An analysis of Siberian watersheds, J. Geophys. Res., 112, F02S22, doi:10.1029/ streamflow hydrology in the Kuparuk River basin, Arctic Alaska: A 2006JF000518. nested watershed approach, J. Hydrol., 206, 39–57. Ye, B., D. Yang, and D. L. Kane (2003), Changes in Lena River streamflow Nelson, F. E., and S. I. Outcalt (1983), A frost index number for spatial hydrology: Human impacts versus natural variations, Water Resour. Res., prediction of ground-frost zones, in Permafrost-Fourth International 39(7), 1200, doi:10.1029/2003WR001991. Conference Proceedings,vol.I,pp.907–911,Natl.Acad.Press, Zhang, T., R. G. Barry, K. Knowles, J. A. Heginbottom, and J. Brown Washington, D. C. (1999), Statistics and characteristics of permafrost and ground-ice distri- Nelson, F. E., and S. I. Outcalt (1987), A computational method for pre- bution in the Northern Hemisphere, Polar Geogr., 23(2), 132–154. diction and regionalization of permafrost, Arct. Alp. Res., 19, 279–288. Zhang, T., O. W. Frauenfeld, R. G. Barry, A. Etinger, J. McCreight, and Pavlov, A. V. (1994), Current changes of climate and permafrost in the D. Gilichinksy (2005a), Response of changes in active layer and perma- Arctic and sub-arctic of Russia, Permafr. Periglac. Proc., 5, 101–110. frost conditions to hydrological cycle in the Russian Arctic, Eos Trans., Peterson, B. J., R. M. Holmes, J. W. McClelland, C. J. Vorosmarty, R. B. AGU, 86, Fall Meet. Suppl., Abstract C21C-1118. Lammers, A. I. Shiklomanov, I. A. Shiklomanov, and S. Rahmstorf Zhang, T., et al. (2005b), Spatial and temporal variability in active layer (2002), Increasing river discharge to the Arctic Ocean, Science, 298, thickness over the Russian Arctic drainage basin, J. Geophys. Res., 110, 2171–2173. D16101, doi:10.1029/2004JD005642. Romanovsky, V. E., T. S. Sazonova, V. T. Balobaev, N. I. Shender, and D. O. Sergueev (2007), Past and recent changes in permafrost and air temperatures in Eastern Siberia, Global Planet. Change, 56, 399–413. D. L. Kane, Water and Environment Research Center, University of Shiklomanov, I. A., A. I. Shiklomanov, R. B. Lammers, B. J. Peterson, and Alaska Fairbanks, 306 Tanana Loop, Fairbanks, AK 99775-5860, USA. C. J. Vorosmarty (2000), The dynamics of river water inflow to the Arctic ([email protected]) Ocean, in The Freshwater Budget of the Arctic Ocean, edited by E. L. D. Yang, Water and Environmental Research Center, Civil and Lewis et al., pp. 281–296, Springer, New York. Environmental Engineering Department, University of Alaska Fairbanks, Smith, L. C., T. M. Pavelsky, G. M. MacDonald, A. I. Shiklomanov, and 306 Tanana Loop, Fairbanks, AK 99775-5860, USA. ([email protected]) R. B. Lammers (2007), Rising minimum daily flows in northern Eurasian B. Ye, States Key Laboratory of Cryospheric Sciences, Cold and Arid rivers: A growing influence of groundwater in the high-latitude hydro- Regions Environmental and Engineering Research Institute, Chinese logic cycle, J. Geophys. Res., 112, G04S47, doi:10.1029/2006JG000327. Academy of Sciences, 320 Donggang West Road, Lanzhou 730000, China. Walvoord, M. A., and R. G. Striegl (2007), Increased groundwater to ([email protected]) stream discharge from permafrost thawing in the Yukon River basin: Z. Zhang, Department of Resources and Environment, Lanzhou Potential impacts on lateral export of carbon and nitrogen, Geophys. University, 222 Tianshui South Road, Lanzhou 730000, China. Res. Lett., 34, L12402, doi:10.1029/2007GL030216. ([email protected]) Woo, M.-K. (1986), Permafrost hydrology in North America, Atmos. Ocean, 24(3), 201–234.

12 of 12