1 Introduction

River discharge is an important element of freshwater budget for the Arctic Ocean and the high-latitude seas. The amount and variation of this freshwater inflow critically affect the salinity and sea ice formation, and may also exert significant control over global ocean circulation (Aagaard and Carmack 1989). Snow cover is a main component of global cryosphere system. Snow cover significantly affects atmosphere, hydrology, permafrost, and ecosystem in the high-latitude regions. Snow cover melt and associated floods are the most important hydrologic event of the year in the northern river basins (Woo 1986; Kane et al. 2000). Recent investigations document snowmelt has started earlier over the recent decades in the northern regions, such as Canada, Alaska, and Siberia, associated with warming in winter and spring seasons (Brabets et al. 2000; Serreze et al. 2000; Whitfield and Cannon 2000; Lammers et al. 2001; Zhang et al. 1999; Ye et al. 2003; Yang et al. 2002, 2014a, b). Studies also demonstrate that the timing and magnitude of northern river streamflow are strongly allied with cold season snow cover storage and subsequent melt (Cao et al. 2002; Yang et al. 2003, 2007). The changes in snowmelt runoff pattern may indicate a hydrologic regime shift over the high latitudes (Serreze et al. 2002; Yang et al. 2002, 2014a, b).

Snow depth data have been routinely collected at the operational networks in the United States and Canada. As a result of differences and changes in methods of observations and data-processing procedures, these data are subject to uncertainties and inconsistency over time and space, particularly across national borders. The operational networks in the northern regions are very sparse. It is therefore a challenge to combine regional snow data to generate basin snow information or gridded products for the high-latitude regions or large watersheds (Robinson 1989; Dyer and Mote 2006). Dyer (2008) used snow depth and discharge data to analyze patterns of snow volume and discharge in major North American watersheds (including the Yukon basin), and found, through statistical analysis, that snow accumulation during late fall and winter is useful to predict spring discharge particularly in the cold Yukon basin and Mackenzie watersheds. Dyer (2008) pointed out that Snow Water Equivalent (SWE) and snow density data should be used to better define the snow accumulation and melt processes; however, reliable SWE and density estimates are difficult to obtain, especially for the continental-scale watersheds in the northern regions. Our knowledge of large-scale snowmelt processes and their interaction with climatic change and variation is incomplete, particularly for the northern regions with insufficient ground-based observations. This limits our ability of understanding past changes and predicting future characteristics of the hydrology system under a warming climate in the high-latitude regions.

Remotely sensed snow data have been very useful to cold region climate and hydrology investigations. For instance, the NOAA weekly snow cover dataset (maps) over the Northern Hemisphere permits quantitative assessments of changes and variations in regional snow extent (Robinson et al. 1993; Serreze et al. 1993; Clark et al. 1997; Frei and Robinson 1999b; Robinson and Frei 2000), and they are useful for hydrologic and snowmelt runoff models (Rango 1996, 1997; Rango and Shalaby 1999). Yang et al. (2003) used the weekly NOAA Snow Cover Extent (SCE) data to study streamflow hydrology in the large Siberian rivers, and discovered that SCE could predict spring discharge with the acceptable accuracy. In addition, long-term SWE data have been derived from the passive microwave sensors (Chang et al. 1987; Chang 1997; Armstrong and Brodzik 2001, 2002). Their potential utility for large-scale hydrology and climate studies in the high-latitude regions has not been fully evaluated. This chapter assesses the compatibility of the passive microwave SWE data over the Yukon watershed, and examines the streamflow response to snow cover change particularly during the spring melt season. The objective is to determine the potential of using remotely sensed snow cover information to enhance our capability of snowmelt runoff modeling over the large northern regions with continuous and discontinuous permafrost. Changes in seasonal snow cover conditions may have significantly contributed to the ground surface temperature increase. The influence of seasonal snow cover on soil temperature, soil freezing, thawing processes, and permafrost has considerable impact on carbon exchange between the atmosphere and the ground and on the hydrologic cycle in cold regions/cold seasons (Zhang 2005). The methods and results of this analysis should improve our understanding of hydrologic effects of a shrinking cryosphere.

2 Basin, Datasets, and Methods

The Yukon has a drainage area of 840 000 km2. It is 3185-km long, with 1149 km in Canada. The Yukon River rises from the Atlin Lake and flows northwest to Fort Yukon, and then turns southwest and enters the Bering Sea (Fig. 9.1). Its major tributaries are the Teslin, Pelly, White, Stewart, Porcupine, Tanana, and Koyukuk rivers. The Yukon basin consists of 37% rolling topography and gentle slopes, 24% low mountains 20% plains and low mountains, 17% moderately high rugged mountains, and 2% extremely high rugged mountains. The Yukon River is one of the largest rivers in the northern regions. It contributes 203 km3 year−1 freshwater to the Bering Sea. It is the fifth-largest river in terms of annual total discharge in the northern regions. Hydrologic conditions and its changes in the Yukon River significantly affect regional biological and ecological systems. The US Geological Survey and Environment Canada maintain a hydrologic network in the Yukon River basin. In this study, long-term daily discharge records collected at the basin outlet station (the Pilot station, 61°56′10″N 162°53′0″W, near the river mouth) during 1975–2001 are used for analyses. Large dams and reservoirs were built in the northern regions for power generation, flood control, and irrigation. Studies show that reservoirs’ regulation alters hydrologic regimes particularly in the regulated sub-basins (Ye et al. 2003; Yang et al. 2004a, b). There are no large dams in the Yukon basin; discharge data collected over this basin are reliable indicators of climate change and variation. The USGS produced a report to document the major hydrologic patterns within the basin (Brabets et al. 2000). Ge et al. (2012) examined Yukon basin hydrologic and climatic changes and variations. Yang et al. (2014b) calculated heat flux for the Yukon River.

Fig. 9.1
figure 1

The Yukon River basin and its tributaries

Full size image

Maps of snow extent and SWE derived from passive microwave satellite data (Scanning Multichannel Microwave Radiometer (SMMR) and Special Sensor Microwave Radiometer (SSMI)) for the Northern Hemisphere have been produced at the National Snow and Ice Data Center (NSIDC) (Armstrong and Brodzik 2001, 2002) using a modified version of the Chang et al. (1987) algorithm. The validation data set used in the Armstrong and Brodzik (2002) study was a topographically consistent subset of data from the ‘Former Soviet Union Hydrological Surveys’ (FSUHS) (Haggerty and Armstrong 1996). This subset (45–60 north latitude 25–45 east longitude) contains a high station density (approximately one transect per 100-km grid cell) and is primarily composed of non-complex terrain (grassland steppe) with maximum elevation differences of less than 500 m. The validation results indicated a general tendency for the algorithms to underestimate SWE in the range of 5–25 mm, particularly as forest cover density begins to exceed 30–40%. Regional maps and products have also been developed in Canada from the SMMR and SSM/I data, and used for analyses of snow cover variations over space and time (Walker and Goodison 1993; Derksen et al. 2000; Walker and Silis 2002). The algorithm by Armstrong and Brodzik (2002) is not able to consistently detect wet snow, only night time or early morning (‘cold’) orbits are used in most analysis, so as to reduce the chance that wet snow is present. Oelke et al. (2003) applied these SWE data for the active layer depth modeling in the Arctic and produced reasonable results. This analysis derived and applied daily SWE data for the Arctic watersheds as part of the effort to examine the large-scale seasonal and inter-annual variations of snow cover and its linkage with river flows (Yang et al. 2003, 2007).

In addition, the NOAA weekly snow cover maps based on visible data are quite reliable to map Snow Cover Extent (SCE) at many times and in many regions including the high latitudes. Mapping frequency and spatial resolution increased with the introduction of daily Interactive Multisensor Snow and Ice Mapping System (IMS) maps in June 1999. However, a lower resolution weekly map has continued to be generated from the IMS and is used in this analysis (Robinson 2003). This pseudo-weekly map involved taking high-resolution IMS grid cells for the fifth map of a week (continuing the weekly NOAA calendar) and determining whether more than 38% (determined from a comparison of both products that were produced independently during 2-year evaluation period) of the 64 IMS cells within a coarse resolution weekly cell were snow covered. If so, the coarse cell was considered snow covered (the 38% value was). Intercomparisons of visible, microwave, and station data derived weekly snow maps suggest strong agreement between the three, though admittedly lower in mountainous regions and near the periphery of the snowpack. The SCE maps have been widely used for hydrologic and climatic analyses in the cold regions, such as development of basin snow cover depletion curves (Rango 1996, 1997), study of streamflow response to snow changes in large northern rivers (Yang et al. 2003), input snow cover data to regional hydrologic and snowmelt runoff models (Rango 1997), and validation of climate model performance (Frei and Robinson 1999a; Yang et al. 1999).

This analysis used the daily EASE-Grid brightness temperature data from NSIDC (nsidc.org/data/nsidc-0032) to run the SWE algorithm (Armstrong and Brodzik 2001, 2002; Brodzik and Knowles 2002), for cold passes without the vegetation corrections, and produced daily SWE data for the northern regions including the Yukon watershed. The basin-mean SWE time-series have been generated from the daily records by averaging all pixels in the watershed. The weekly data have been generated by averaging the 7 day basin-mean SWE record during 1988–2001. On the basis of these weekly records, we examine the seasonal changes of snow cover mass, by defining the SWE climatology based on weekly statistics, determining the dates of snow cover formation/disappearance and duration of snow cover/snow-free days, and quantifying the rates of snow cover mass change during the accumulation and melt seasons. We also derive weekly discharge time-series from the daily streamflow data collected near the basin outlet, and use the weekly data to describe the seasonal streamflow characteristics, including discharge regime, rates of streamflow rise, and peak flow during the melt period. We calculate the weekly correlation of streamflow with basin SWE, and determine the consistency between SWE and streamflow changes over the seasons. Furthermore, we identify extreme snowmelt streamflow cases and examine their correspondence with basin snow cover conditions. These analyses characterize the weekly relationship between snowmelt runoff and basin SWE changes for the Yukon River. In addition to streamflow and snow cover data, basin-mean weekly precipitation and temperature time-series during 1966–1998 have been created based on gridded global data sets (Jones 1994), and used to investigate the compatibility of SWE/SCE data with climate variables and to explain the streamflow response to seasonal snow cover changes.

It is important to note that the approach of this analysis is not a complete water budget calculation; rather, we focus on the major terms in basin water budget, i.e., SWE, winter precipitation, and streamflow. We relate snow cover data (SWE and SCE) with streamflow data measured near the basin outlet, since discharge represents the integrated response of basin hydrology to climate influence. To be compatible with discharge data, we need basin-mean snow and climate data for our analysis. To generate basin-averaged data, it is necessary to define the basin boundary. A river network grid by Fekete et al. (2001) was used over the Yukon basin and overpaid onto the gridded snow and climate data. All the grids inside the basin and those with more than 50% within the basin boundary were counted as the basin grids and used to produce basin averages. A simple average was calculated without taking into account topography effects on snow, precipitation, and temperature distributions. Similar approach has been applied for other large watersheds in North America and Siberia (Yang et al. 2003; Dyer 2008). Given the large size of the Yukon watershed and the focus of this analysis on the consistency examination of various data, the use of basin average is effective and appropriate.

3 Basin Hydro-Climatic Regime

To understand the climatic and hydrologic regime of the Yukon basin, we present the basin-mean temperature, precipitation, snow cover, and discharge data. The weekly timescale is used for accurate discussions of the seasonal cycles of basin hydrology and climate, particularly for snow cover and streamflow. Figure 9.2 shows the basin-mean weekly temperature regime. The basin is cold, with mean temperatures between −10 and −20 °C, during weeks 42–14; and it is relatively warm during weeks 20–38, when mean temperatures vary from 0 to 12 °C. Basin temperatures rise from −10 °C to near 0 °C during weeks 15–19, and decline from 0 to −10 °C during weeks 39–41. The inter-annual fluctuations in winter temperatures are significantly greater (ranging from −35 to −5 °C) than the differences in the summer temperatures (ranging from 8 to 15 °C). During the period 1966–2002, Yukon basin-mean annual air temperature is about −5.1 °C (SD = 0.98 °C), with the lowest annual air temperature of −6.8 °C in 1974, and the highest being −2.9 °C in 1993. Similar to other northern regions, Yukon basin has experienced a significant (α = 0.05 level) warming trend (0.03 °C/year) during 1966–2002.

Fig. 9.2
figure 2

Basin-mean weekly temperature in (T, °C) for the Yukon watershed, 1966–2002

Full size image

Similar to temperatures, a strong seasonal cycle exists for precipitation over the Yukon basin (Fig. 9.3). Precipitation is high (10–15 mm) during the warm season (weeks 21–43) and low (5–10 mm) during the winter season (weeks 44–20). During the period of 1988–2002 (n = 15), the mean annual total precipitation in Yukon basin is 384.6 mm (SD = 37.7 mm). The lowest annual precipitation is 312.3 mm in 1999, and the highest annual precipitation is 442.0 mm in 1994. The mean annual total snowfall (defined as the precipitation during the period when basin-mean air temperature is below 0 °C) is 180.4 mm (or 47% of the yearly total precipitation). The lowest annual snowfall is 102.3 mm in 1998, and the highest annual snowfall is 258.9 mm in 1992. Winter precipitation is generally lower than summer precipitation over the Yukon basin. On the 52nd week of 1999 and the 3rd week of 2000, weekly precipitation was about 31 and 24 mm, respectively; they were unusually higher for winter season perhaps due to strong snowfall events over the lower parts of the basin. No significant trends were found for yearly total precipitation and annual total snowfall for the study period 1988–2002.

Fig. 9.3
figure 3

Basin-mean weekly precipitation (P, mm) over the Yukon watershed, 1988–2002

Full size image

Figures 9.4 and 9.5 show the basin weekly Snow Cover Percent (SCP) and SWE cycles. They illustrate snow accumulation during weeks 37–44, when basin SCE rises from 10 to 90%; a complete snow cover (100% SCE) during weeks 44–16; snowmelt during weeks 17–24, when SCE drops from 90 to 10% over the basin; and a minimum SCE of 5–10% in summer (weeks 25–36) due to glacier and snow cover in the high elevations. The SCE data show significant inter-annual variations. For the maximum SCE case, we see 100% SCE lasting during weeks 40–24, a fast melt in 2–3 weeks during weeks 25–28, a higher minimum SCE of 20–30% in the mid-summer, and an early snow accumulation during weeks 36–40. On the other hand, for the minimum SCE case, we observe a 100% snow cover during the weeks 50–16, a slow snowmelt in weeks 17–22, almost snow-free during weeks 24–39, and a late snow accumulation in weeks 40–48.

Fig. 9.4
figure 4

Basin-mean weekly snow cover percentage (SCP, %) over the Yukon watershed, 1966–2002

Full size image
Fig. 9.5
figure 5

Basin-mean weekly snow water equivalent (SWE, mm) over the Yukon watershed, 1988–2001

Full size image

Similar pattern exists for the SWE (Fig. 9.5). Snowpack accumulates during weeks 38–4, reaches a stable state during weeks 5–16, with the mean max SWE of 90–100 mm around weeks 8–12. Snowpack starts to shrink slightly after the peak SWE. Snow cover melts during weeks 16–22, with SWE dropping from the peak to near 0 mm at week 23. Basin becomes snow-free during weeks 23–37, when the average SCE is about 5%, with max being as high as 30%, due to glaciers and snow cover in the high elevations. SWE condition varies among the years, for instance, the peak SWE range from 60–70 to 110–120 mm. The rate and length of snowmelt season also varies greatly, depending on the spring temperature and amount of peak SWE. A shorter melt season usually suggests a faster melt of snow cover due to late onset of melt associated with higher temperatures during late spring. During the period of 1988–2002, the mean peak SWE in Yukon basin is about 106 mm (SD = 130 mm). The lowest and highest peak SWEs are 81 mm in 1991 and 125 mm in 1998, respectively. No significant trends were found for the SCP or SWE over the study period 1988–2002.

The seasonal cycle of discharge near the basin outlet is illustrated in Fig. 9.6. It generally shows low flows during November–April (weeks 45–17), highest flow in June (weeks 22–24) due to snowmelt runoff, high flow in summer (weeks 25–40) due to glacier melt, and recession of flow in fall season (weeks 40–44). Streamflow of the Yukon River peaks at weeks 22–24 (or mid-June), when the basin is covered by a small patchy snowpack, i.e., approximately 4% SCE and 1 mm SWE over the watershed. The basin SWE amounts are very low at the time of peak streamflow, perhaps reflecting a long lag of streamflow response to snowmelt and flow routing within the large watershed. On the other hand, it should be noted that the SWE algorithm is more appropriate for the temperature gradient typical of a deep and dry snowpack. SWE will decrease in the presence of even small amount of liquid water, as soon as the cold passes start observing liquid water, since the emission from the water effectively reduces the temperature gradient. The SWE may fall off much faster than the real melt rate.

Fig. 9.6
figure 6

Mean weekly discharge (m3/s) near the basin outlet, 1975–1996

Full size image

Streamflow decreases at the end of the snowmelt season, although heavy rainfall events and glacier melt in mid-summer generate secondary floods over the basin. The inter-annual variations of weekly streamflow are generally small in the cold season and large over summer months mainly due to rainfall storm activities and associated streamflow fluctuations. It also indicates noticeable differences in streamflow characteristics between years mainly due to different climate and snow conditions over the basin. The annual peak discharge in Yukon basin is 19 178 m3/s, with the lowest and highest peak discharge being 12 969 m3/s in 1978, and 30 299 m3/s in 1985, respectively. It is important to note that the year having the highest annual flow was the same year for the highest peak flow, and the year with the lowest annual flow also has the lowest peak flow. This indicates that the peak flow in spring and summer dominates the annual flow. No significant changes were found for peak discharge or annual total discharge during the study period.

4 Compatibility of Basin Snow Cover Data

Temperature and precipitation are the main factors affecting snow cover characteristics including accumulation and melt processes. To understand the winter snow-mass budget, we compare basin SWE with the Accumulated Precipitation (AP) over the period when the basin-mean weekly temperatures are below 0 °C. The AP may include some rainfall events in early spring and late fall seasons particularly over the southern parts of the watersheds. The contribution of rainfall events is small to the winter total precipitation.

Figure 9.7 shows significant variation in the AP among the years. Both the high and low AP winters were associated with similar SWE amounts, although in most years basin SWE is generally less than the AP during the snow cover season. It is interesting to note that the amounts of maximum SWE are closer to AP in low snowfall winters, and much less than AP in high snowfall winters. These results are consistent with the saturation in the temperature gradient with the modeling work by Chang et al. (1987). It is reasonable to expect that basin maximum SWE should be generally close to winter total snowfall amount. The lack of correspondence of the basin SWE to AP variation indicates some inconsistency between the SWE and precipitation data. This is not completely unexpected given the large biases in snowfall data over the high-latitude regions (Yang et al. 1998, 2005; Yang and Ohata 2001) and limitations in remote sensing snow cover algorithm, including the issue of saturation (Walker and Goodison 1993; Armstrong and Brodzik 2001). In addition, sublimation loss from snowpack over winter is another factor contributing to the uncertainty in SWE and AP compatibility. As discussed in Chaps. 3 and 4, many studies reported that sublimation from the snow surface accounts for up to 1/3 of total accumulation in the northern regions (Benson 1982; Liston and Sturm 1998, 2004). Sublimation over large basins and regions is difficult to determine through direct measurements. Snow models taking into account the blowing and drifting snow processes can provide reasonable estimate of regional winter sublimation amount (Liston and Sturm 1998, 2002). The ratios of maximum basin SWE versus AP are 37–120% (mean 77%) for the Yukon River. The ratios close to 100% reflect less difference between the SWE and AP.

Fig. 9.7
figure 7

Comparisons of basin snow water equivalent (SWE, mm) with winter Accumulated Precipitation (AP, mm) from 1988/1989 to 2000/2001 (Zhao 2004)

Full size image

The inter-annual variations in the SWE/AP ratios are mainly due to fluctuations in snowfall amounts and temperatures over the winter season. The low (high) ratios are found associated with high (low) AP and warm (cold) winter. It is important to note that the Yukon basin SWE was greater than the AP for several winters (i.e., 1995–1996, 1996–1997, 1997–1998, and 1998–1999). This unexpected result indicates uncertainties in the SWE and AP estimations over the region. There might be possible SWE algorithm saturation, as the basin-average SWE never exceeded 100 mm, regardless of winter precipitation variation. In addition, precipitation gauge undercatch of snowfall is also a factor, since studies (Yang et al. 1998, 2005; Benning and Yang 2005) found underestimation of yearly precipitation by 25–50% over Alaska. In addition, determination of timing of snow cover accumulation is also a challenge. In this analysis, basin-mean temperatures at 0 C were used to estimate the beginning date (week) of snow cover formation, i.e., the starting point for AP. Given the very large size of the watershed, basin-mean temperatures do not always represent the thermal conditions over the entire basin, particularly during spring and fall transition periods. Sub-basin scale analyses might be necessary to better examine the compatibility between basin SWE and winter precipitation.

To examine and quantify the impact of temperature on basin snow cover conditions, a linear regression is applied to the temperature and SCE/SWE data sets for each week in a year. Figures 9.8 and 9.9 present the scatter plots of (snow cover percent or SCE) SCP versus temperature and SWE versus temperature for the 53 weeks. The regression functions and R2 are displayed in the plots for the weeks with significant relationships. The results generally show that the snow cover changes as a function of temperature. The basin SCP and SWE are the highest in the beginning of the year when temperatures are very cold between −15 and −20 C. Both SCP and SWE decrease in spring from very high to very low in a short time period when basin-mean temperatures are around 0 C. The basin is almost snow-free in the short summer season except in the mountain regions with glaciers and snow cover all year around. Snow cover forms when temperatures drop back to around 0 C in fall and continues to accumulate over the fall–winter seasons. Regression analyses identify strong negative correlations between basin SCP/SWE and temperature, particularly when temperatures are close to 0 C during the snow accumulation and melt seasons (Figs. 9.8 and 9.9). These correlations demonstrate the association of (high) low SWE/SCP with (low) high temperatures.

Fig. 9.8
figure 8

Scatter plots and regression equations of basin weekly SCP (%) versus weekly air temperature (°C) for the 53 weeks in a year during 1966–2001 (Zhao 2004)

Full size image
Fig. 9.9
figure 9

Scatter plots and regression equations of basin weekly SWE (mm) versus weekly air temperature (°C) for the 53 weeks in a year during 1988–2001 (Zhao 2004)

Full size image

5 Weekly Relation Between Streamflow and Basin SWE

The seasonal changes of the basin SWE and streamflow in each individual year are displayed in Fig. 9.10. They clearly indicate a general response of river runoff to seasonal snow cover changes, i.e., an association of high discharge with low SCE and SWE during summer, a decrease in discharge associated with increasing SCE and SWE in fall, an association of low streamflow with high SCE and SWE during the cold season, and an increase in discharge associated with decreasing SCE and SWE during the spring melt periods. They also show the inter-annual variations in both SWE and streamflow. Relative to the basin SWE, streamflow varies much more between years. For instance, the Yukon River peak streamflow was high (23 000 m3/s) in 1991 and low (18 000 m3/s) in 1996, while the maximum basin SWE was low (about 75 mm) in 1991 and high (about 90 mm) in 1996. A similar peak SWE was found for 1988, 1989, 1990, 1994, 1995, and 2001, while the spring peak flow differs significantly among these years, particularly between 1994 (low flow) and 2001 (high flow). This discrepancy between basin snow cover and streamflow variations may suggest uncertainties in basin SWE data perhaps due to algorithm limitations (Armstrong and Brodzik 2002), particularly for the mountain regions within the Yukon basin.

Fig. 9.10
figure 10

Comparisons of basin Snow Water Equivalent (SWE, mm) with river discharge (Q, m3/s) during 1988–1996 and 2001 (Zhao 2004)

Full size image

To quantify the response of river streamflow to basin snow cover variation, we examine and compare the weekly mean streamflow with the weekly basin SWE for the study period 1988–1999. The results generally confirm a meaningful relationship between the streamflow and SWE during the spring melt season over the Yukon watershed (Fig. 9.11). In the early melt period (weeks 16–18), basin SWE reduces from 120 to 50 mm. Most of the meltwater is stored in ponds, lakes, and river valleys. River ice breaks up around this time in the upper parts of the basin, but streamflow at the basin outlet does not show a clear response due to ice jams in the river valleys. As snowmelt progresses (weeks 19–21), SWE decreases from 70 to 20 mm, releasing more water to satisfy the surface storage within the basin. During weeks 22–24, river channels open up in the downstream parts of the watershed and discharge near the basin mouth starts to rise and reach the maximum. This response of streamflow to snowmelt is reflected by a negative correlation between streamflow and basin SWE in weeks 19–21. In the late melt period (weeks 23–25), streamflow response to snowmelt weakens due to reduced snowmelt runoff contribution. The results of regression analyses are shown in Fig. 9.11. They explain 30–70% of streamflow variability, although they are statistically significant at 85–95% confidence. It is useful to derive these relationships, as they suggest a practical procedure of using remotely sensed SWE information for snowmelt runoff estimation over the large northern watersheds.

Fig. 9.11
figure 11

Scatter plots and regression equations of weekly discharge (Q, m3/s) versus weekly basin Snow Water Equivalent (SWE, mm) for the 53 weeks in a year, 1988–2001 (Zhao 2004)

Full size image

6 Extreme Streamflow and Associated Snow Condition

The basin snow cover and discharge data show that weekly snow cover and snowmelt peak flows vary significantly among years. To better understand the variability in snowmelt runoff, it is useful to examine extreme streamflow and its association with the snow cover condition, such as the peak accumulation and melt process. Two sample years of highest and lowest streamflow cases were selected, i.e., peak flows during the snowmelt season of 13 900 m3/s in 1996 and 24 110 m3/s in 1990. Figure 9.12 compares the SCE and SWE data with the extreme streamflow for the two years. It shows that, for the SCE, a similar process (rate) of snow cover depletion during the early melt season (weeks 17–20) between the two years, and a slower melt and longer SCE recession for the low flow year during the late melt season (weeks 21–25) (Fig. 9.12a). The peak snow accumulation over the basin is about 115 mm for the high flow year of 1990 and 100 mm for the low flow year of 1996, indicating higher (lower) flows associated with higher (lower) basin SWE. The melt patterns were very similar between the two extreme years, with the snowmelt beginning around week 16 and ending around week 21. The timing of the peak flow was 1 week earlier in the low flow year than the high flow year; the shape of the spring hydrograph is sharp (with a single high peak) in the high flow year and flat (with two low peaks) in the low flow year. It is important to note that the difference in peak SWE was only 15 mm, while the peak spring flow difference was about 10 210 m3/s (Fig. 9.12b). The difference in peak flows is much higher than the difference in basin SWE between the extreme years. This seems to suggest inconsistency between basin SWE and streamflow. In addition to the winter maximum SWE over the basin, other factors such as temperature, precipitation, soil moisture condition, and ground thaw processes during the melt periods also affect snowmelt processes and influence the timing and magnitude of peak snowmelt floods.

Fig. 9.12
figure 12

Comparison of extreme discharge (Q, m3/s) cases with SCP (a) and SWE, (b) over the Yukon watershed (Zhao 2004)

Full size image

7 Summary and Conclusion

Validation and evaluation of available remotely sensing products are important to develop our capability of observing and monitoring the earth system from the space. This analysis applied remotely sensed SWE, SCE, and gridded climatic data to investigate snowmelt runoff response to seasonal snow cover change over the Yukon watershed. It defined the seasonal cycles and variations of snow cover and river streamflow, and identified a clear correspondence of river streamflow to seasonal snow cover change, i.e., an association of low streamflow with high snow cover mass during the cold season and an increase in discharge associated with a decrease of snow cover extent and SWE during the melt periods. It also examined the compatibility of the basin SWE data with the SCE, peak snowmelt floods, and climatic variables (temperature and winter precipitation), and found consistency among the basin SWE, SCE, and temperature. On the other hand, it detected incompatibility between basin SWE and winter precipitation, suggesting limitations in SWE retrieval algorithm and uncertainties in the determination of basin winter snowfall amounts.

To quantify the relation between river streamflow and basin snow cover variations, we compared the weekly mean streamflow with the weekly basin SWE for the study period. The results revealed a meaningful linkage between streamflow and basin SWE during the spring melt season, and developed a statistically significant weekly streamflow–SWE relationship. It is important to explore these relationships, as they improve our understanding of the most important arctic hydrologic process—snowmelt peak floods—and they also suggest practical procedures of using remotely sensed snow cover information for snowmelt runoff forecasting over the large northern watersheds with insufficient ground observations. Furthermore, analyses of extreme (high/low) streamflow cases (years) and basin snow cover conditions indicate a general association of high (low) flow peak with high (low) maximum SWE over the basin, although some inconsistencies exist between extreme flow and basin SWE. These results point to a need to further search for the best snowmelt–streamflow relationship, and to develop the most useful snowmelt runoff forecasting methods for the large northern rivers. There are uncertainty and saturation problems in the SWE algorithms (Clifford 2010). There is good improvement in passive microwave SWE algorithms, including the AMSR-E (Kelly 2009; RSSJ) that includes adjustments for shallow show and a dynamic density model.

The results of this analysis demonstrate that remote sensing snow cover data are useful in understanding streamflow characteristics and changes in the arctic regions with a very sparse observational network. The methods and results of this investigation are important to seasonal hydrologic forecasting, snowmelt model, and process studies. They improve our understanding of the spatial and temporal variability of high-latitude snow cover and its contribution to river runoff in the northern rivers. Snow depth and water equivalent data obtained by ground observations are also useful to better understand snowmelt and runoff processes (Dyer 2008; Brown et al 2010). Long-term snow observations particularly over the Siberian regions have been found valuable for cold region climate studies (Ye et al. 1998; Armstrong and Brodzik 2001). There is a need to investigate the compatibility of the SWE with in situ snow cover observations at sub-basins scales for more detailed analyses of snow–runoff relationship. It is also necessary to integrate remote sensing and ground-based snow datasets for land surface process modeling and simulation of hydrologic change over the cold regions.