close

Вход

Забыли?

вход по аккаунту

?

9780784482346.022

код для вставкиСкачать
World Environmental and Water Resources Congress 2019
On the Role of Spatial Snow Distribution on Alpine Catchment Hydrology
Andrew M. Badger, Ph.D.1; Ben Livneh, Ph.D.2; and Noah P. Molotch, Ph.D.3
Downloaded from ascelibrary.org by University of Liverpool on 08/26/19. Copyright ASCE. For personal use only; all rights reserved.
1
Cooperative Institute for Research in Environmental Science (CIRES), Univ. of Colorado
Boulder, Boulder, CO 80309. E-mail: [email protected]
2
Cooperative Institute for Research in Environmental Science (CIRES) and Dept. of Civil,
Environmental and Architectural Engineering, Univ. of Colorado Boulder, Boulder, CO 80309.
E-mail: [email protected]
3
Dept. of Geography and Center for Water, Earth Science, and Technology, Univ. of Colorado
Boulder, Boulder, CO 80309. E-mail: [email protected]
ABSTRACT
The spatial distribution of alpine snow is largely governed by topographic complexity
through shading of solar radiation and wind redistribution. While radiative shading is relatively
consistent between years, wind redistribution is more variable, capable of producing anomalous
areas of deeper snow that can persist into the late-summer and early-fall months. In a warmer
future climate, we hypothesize that snow will become wetter and heavier than in the current
climate, leading to less wind redistribution and a more homogeneous snowpack. This study seeks
to investigate how the relative degree of snow redistribution will impact total streamflow
generation in an alpine environment. A distributed hydrologic model is applied together with
remotely sensed spatial estimates of snowpack in order to quantify the impact of spatial
variability of snow water equivalent on streamflow. Preliminary findings for the Green Lakes
Valley within the headwaters of the Boulder Creek watershed in Colorado, 2001–2014, show
that more uniform snow distributions lead to earlier melt-out of 47 days on average and tend to
produce less overall streamflow, with maximum decreases as large as 9.5% relative to the natural
case. Overall, this analysis aims to provide insight into how warming-mediated changes in snow
distribution will impact water resources in critical streamflow generating headwaters catchments.
INTRODUCTION
Alpine ecosystems have been described as among the most vulnerable to climate change
(Jones et al., 2012; Field et al., 2014). Sharp physical gradients in these ecosystems mean small
perturbations in temperature can turn ice and snow to liquid water (Scherrer and Körner, 2011,
Suding et al., 2015). Vaughan et al. (2013) further highlights winter snow cover as one of the
fastest changing climate features. In regions that feature snowmelt dominated runoff, like the
Niwot Ridge in Colorado, USA, understanding potential hydroclimatic changes are of utmost
importance.
As part of the Niwot Ridge Long-Term Ecological Research project, a record of
meteorological observations dating back to the 1950s are available, the longest high-elevation
continuous climate record in the US (McGuire et al., 2012). Climate records indicate that Niwot
Ridge is experiencing earlier spring snowmelt, warmer springs and warmer summers (Kittel et
al., 2016; McGuire et al., 2012). Niwot Ridge is not an anomalous location, regionally across
Colorado and the rest of the western U.S., observations show that snowmelt timing is occurring
approximately three weeks earlier since the 1970s (Clow, 2010; Stewart et al., 2005; Mote et al.,
2005; Hamlet et al., 2005; Pierce et al., 2008;). Alteration of this local hydroclimatology is key,
as the impacts of earlier snowmelt alters streamflow generation and antecedent soil moisture
© ASCE
World Environmental and Water Resources Congress 2019
215
Downloaded from ascelibrary.org by University of Liverpool on 08/26/19. Copyright ASCE. For personal use only; all rights reserved.
World Environmental and Water Resources Congress 2019
conditions (Huntington and Niswonger, 2012; Godsey et al., 2014; Stewart et al., 2005).
One potential element leading to hydrologic changes is snow redistribution by wind. Essery
et al. (2004) notes that as temperatures rise, there is a decreased probability of snow
redistribution due to wind. Two mechanisms for this effect are that (i) snow in warmer
temperatures is of higher density and therefore requires more lateral momentum to transport and
hence is less redistributed, and (ii) warming produces less snow and more rain, hence shallower
snow depths for a given quantity of water, thus producing a denser snowpack that is less likely to
get redistributed by wind.
Due to the complex topography of alpine watersheds, high-resolution physically-based
computational modeling of these catchments is still relatively limited, with existing work
primarily describing first-order sensitivities to climate changes. A number of studies have used
models that discretize alpine watersheds into soil-land combinations in semi-physically based
models to assess water quality, quantity and runoff processes (e.g. Abbaspour et al., 2007; Gurtz
et al., 1999). With the recent advent in computational capacity and high-resolution observations,
we can now advance a number of studies using physically-based models at a resolution that was
once too high to be properly realized within the model (e.g. Chet at al., 2014; Graham et al.,
2007; Livneh et al., 2014).
In this study, we will integrate remotely-sensed estimates of spatial snow distribution with
hydrologic modeling, to investigate the impact of alterations to snowpack distributions on
streamflow generation in the Green Lakes Valley, an area within the headwaters of the Boulder
Creek watershed in Colorado, USA. The following sections of the paper will describe the
hydrologic model, region of interest and modelling simulations used in this study. The analysis
of hydrologic model simulation results is followed by a discussion of the implications of
hydrologic changes in a potential future climate.
Figure 1. Map of Niwot Ridge elevation with Green Lakes Valley outlined, and sites of
meteorological stations (Table 1) used to provided forcing data for DHSVM.
METHODS
Study Area: The Niwot Ridge LTER is approximately 35 km west of Boulder, Colorado and
bounded to the west by the Continental Divide, with the entirety of the domain residing above
© ASCE
World Environmental and Water Resources Congress 2019
216
Downloaded from ascelibrary.org by University of Liverpool on 08/26/19. Copyright ASCE. For personal use only; all rights reserved.
World Environmental and Water Resources Congress 2019
3000 m. Within the Niwot Ridge LTER is the Green Lakes Valley (GLV; Figure 1), our specific
alpine catchment of interest for this study. The GLV is 2.3 km2 and spans the elevation range of
3250 m to 3798 m. Williams et al. (1996) notes that annual mean temperature of the region is 3.8°C and receives about 1000 mm of precipitation annually, about 80% of which is snow
accumulating from October to April (Caine, 1996). The snowmelt dominated runoff accounts for
70% of the total annual runoff, peaking between late-April and mid-July (Caine, 1996).
Model Description: The Distributed Hydrology Soil and Vegetation Model (DHSVM;
Wigmosta et al., 1994) consists of a two-layer energy balance model for snow accumulation and
snowmelt, a multilayer unsaturated soil model and a saturated subsurface flow model, and a twolayer canopy representation for evapotranspiration (ET) and energy transfer. DHSVM accounts
for slope and aspect in characterizing incoming radiation (shortwave and longwave) within the
surface energy budget.
Static model layers (i.e. not varying with time) of vegetation, soil depth and texture, geology,
and topography were generated based on local observations from the Niwot Ridge Long Term
Ecological Research site collocated with GLV. Time varying model layers include shading files
for each month, comprised of the mean diurnal cycle for cloud free incoming solar radiation. The
model was set up at a 20 m horizontal resolution and was run at an hourly time-step, 2001-2014.
DHSVM was previously applied over the Boulder Creek watershed (Livneh et al.,
2014;2015)—within which the Green Lakes Valley is a tributary—realistically simulating
snowmelt and streamflow dynamics. The Livneh et al. (2014;2015) model set up is used here to
provide initial settings for soil and vegetation parameters. Additional parameter adjustments
were made following a Monte Carlo search procedure to more closely match gauged streamflow
at the GLV outlet, ultimately leading to a daily Nash-Sutcliffe Efficiency (NSE; Nash and
Sutcliffe, 1970) of 0.604 after calibration.
Meteorological Data: The hourly meteorological time-series information used to drive
DHSVM was derived from five observation locations within the Niwot Ridge domain (see Table
1 and Figure 1). As is a common issue with surface observations in extreme climatic regions,
outages and lost data does occur. The infilling of missing data was done on the basis of
correlation with nearby stations and insertion based on the monthly mean diurnal cycles.
Name
Arikaree
D1
GL4
Saddle
TVan
Table 1. Meteorological sites used in DHSVM simulations.
Latitude (°)
Longitude (°)
Elevation (m)
40.049
-105.640
3798
40.059
-105.616
3743
40.056
-105.617
3560
40.049
-105.592
3525
40.053
-105.586
3480
Meteorological variables are interpolated to all grid cells automatically within DHSVM using
a Cressman scheme that can be informed by additional spatial information to distribute the local
meteorology. In order to construct a realistic baseline distribution of snowpack (Figure 2a), we
use the remotely sensed SWE product of Jepsen et al. (2012) as a co-variant in the interpolation
of station data within the model. The Jepsen et al. (2012) SWE reconstruction provides an
estimate of the yearly maximum SWE by integrating modeled snowmelt from the date of
maximum SWE to the date of observed disappearance using observed meteorology to estimate
energy balance calculations. The 12-year mean of the SWE reconstruction, 1996-2007, was used
© ASCE
World Environmental and Water Resources Congress 2019
217
World Environmental and Water Resources Congress 2019
Downloaded from ascelibrary.org by University of Liverpool on 08/26/19. Copyright ASCE. For personal use only; all rights reserved.
to distribute the meteorological forcing, the time-period mean is chosen due to the lack of
overlapping years in the SWE reconstruction and our model simulation.
Figure 2. Map of mean SWE reconstruction product (a; Jepsen et al., 2012) and an example
of subsequent iterations for smoothing SWE distributions (b-g) used in sensitivity
simulations.
While the SWE reconstruction inherently has wind redistribution of the snowpack accounted
for, inclusion of the product allows for more realistic spatial variability in the distribution of the
meteorological forcing. Livneh et al., (2015) found that the integrating the SWE reconstruction
data within DHSVM helped improve the simulated water balance.
Model Simulations: The model is run continuously for the full time- water-years 2001-2014,
and model states are saved for April 1 of each year. April 1 is chosen as a common reference
date because it is commonly used as the primary date for summer runoff generation prediction in
the region.
For each saved state, six variations of SWE distributions were developed (Figures 2b-g) with
varying degrees of smoothing. For all scenarios, the total basin-wide SWE is conserved. The first
scenario redistributes 1/6th of the snow mass from locations above the basin SWE mean to
locations below the basin SWE mean. Each subsequent permutation was generated by moving an
addition 1/6th of snow mass and ultimately providing a uniform snowpack of the mean basin
SWE mean in the final iteration. By moving the snow from areas with more snow to less snow,
we are able to undo the wind redistribution of snow that was observed.
For each year, the control and altered snowpack distributions initialized on April 1, are
forced from April 1 through September 30 (i.e. the end of the water-year) with the observed
meteorological forcing for their respective water-year. This experimental design provides an
ensemble of hydrologic model simulations initialized with the control and altered snowpacks for
each year.
RESULTS
The first item to note is that each year provides a different SWE initial condition and
subsequent precipitation for the remainder of the water-year (Figure 3). This interannual
© ASCE
World Environmental and Water Resources Congress 2019
218
Downloaded from ascelibrary.org by University of Liverpool on 08/26/19. Copyright ASCE. For personal use only; all rights reserved.
World Environmental and Water Resources Congress 2019
variability ultimately leads to interannual variability in streamflow generation from the Green
Lakes Valley. This is an artifact of the natural system that can have impacts on the processes
governing streamflow generation on a year-to-year basis.
The simulation of the snowpack for each given distribution (Figure 4) highlights a few
notable findings. The first is that after peak SWE the more uniform snowpacks decline more
quickly. The melt-out date in the control simulation does not occur on average for 170 days
(September 17). While the uniform snowpack persists on average for only 123 days (August 1),
47 days earlier than the control simulation. The melt-out date is sequentially earlier for
increasing uniformity relative to the control simulation. However, the response is non-linear;
when moving half the snow mass, the snow pack persists for 165 days, just 5 days less than the
control simulation. After moving more than half of the snow mass, the snowpack persistence
declines rapidly.
Simulated streamflow generation across the ensembles (Figure 5) shows largest sensitivity
during the period of rapid snowmelt, primarily increased early-season streamflow generation
from more uniform snowpacks with ultimately less cumulative streamflow generation from these
more uniform snowpacks by the end of the water-year. Although early-season streamflow
generation is present when reducing snowpack variability, the date of peak streamflow discharge
is largely unchanged from the control to mean snowpack distributions, with the mean snowpack
peaking just 2 days earlier than control simulation. The uniform distribution produces 2.2% less
streamflow than the control snowpack distribution on average.
Figure 3. April 1 SWE (grey) and April to September precipitation (blue) for each year;
dashed lines represent the respective means for the given time period.
The effect of snowpack uniformity in reducing streamflow appears to be dependent on the
magnitude of snowpack for a given year. Figure 6 shows each year’s initial mean snowpack
values (e.g. April 1 areal mean SWE) and streamflow generation normalized by the control for
each simulation with two key features emerging. First, snowpack uniformity in lower SWE years
tend to produce larger decreases in streamflow generation, that is years with an areal mean SWE
value less than the climatological areal mean (0.496m) tend to show larger decreases in
streamflow generation as snowpack variability decreases, with some decreases by as much as
9.5%. In contrast to the first point, the second feature is that snowpack uniformity in deeper
snowpacks tends to produce more streamflow. We offer two hypotheses for this. One possible
reason is that in such an energy limited climate regime, more incoming energy is partitioned to
© ASCE
World Environmental and Water Resources Congress 2019
219
World Environmental and Water Resources Congress 2019
Downloaded from ascelibrary.org by University of Liverpool on 08/26/19. Copyright ASCE. For personal use only; all rights reserved.
snowmelt and increased runoff efficiency occurs due to early-season soil saturation. A second
possible reason is that subsequent delivery of precipitation (e.g. amount, duration, intensity) is
altering the manner in which the snowpacks are producing streamflow.
Figure 4. DHSVM simulated daily-mean SWE for the control simulation and sensitivity
iterations from April 1 to September 30 across all years. Vertical dashed lines represent
the mean day of melt-out (or minimum) SWE for each respective simulation.
Figure 5. DHSVM simulated daily-mean cumulative streamflow from the Green Lake
Valley for the control simulation and sensitivity iterations from April 1 to September 30
across all years. Vertical dashed lines represent the day of peak streamflow for each
respective simulation.
DISCUSSION
To further investigate the findings of Figure 6, we attempt to isolate the connection between
© ASCE
World Environmental and Water Resources Congress 2019
220
World Environmental and Water Resources Congress 2019
Downloaded from ascelibrary.org by University of Liverpool on 08/26/19. Copyright ASCE. For personal use only; all rights reserved.
snowpack uniformity and snowmelt from confounding factors like the variability in melt-season
meteorology. A new set of simulations were performed in an identical manner to the initial
simulations, except precipitation forcings were removed (i.e. set to zero).
Figure 6. Annual April 1 SWE (horizontal axis) and total annual streamflow generation
normalized by respective control simulations (vertical axis) for all year and each snowpack
distribution. Solid black line indicates a ratio of 1, with dots above (below) the line
indicating an increase (decrease) in streamflow generation.
Figure 7. DHSVM simulated daily-mean SWE without precipitation forcing for the control
simulation and sensitivity iterations from April 1 to September 30 across all years. Vertical
dashed lines represent the mean day of melt-out (or minimum) SWE for each respective
simulation.
Consistent with Figure 4, the removal of precipitation forcing in Figure 7 shows a similar
decline in SWE with increasing snowpack uniformity. Notably, differences in the melt-out date
are more pronounced for the new simulations. A difference of 72 days between the control and
© ASCE
World Environmental and Water Resources Congress 2019
221
World Environmental and Water Resources Congress 2019
Downloaded from ascelibrary.org by University of Liverpool on 08/26/19. Copyright ASCE. For personal use only; all rights reserved.
uniform cases, 25 days more than when precipitation forcing was included. Interestingly, the
control snowpack distribution melt-out date is largely unaffected by the exclusion of
precipitation, yet each of the iterations see a melt-out date occurring 2 to 23 days earlier. This
agreement strengthens the argument that more uniform snowpacks will melt out earlier.
Figure 8. DHSVM simulated daily-mean cumulative streamflow without precipitation
forcing from the Green Lake Valley for the control simulation and sensitivity iterations
from April 1 to September 30 across all years. Vertical dashed lines represent the day of
peak streamflow for each respective simulation.
Figure 9. Annual April 1 SWE (horizontal axis) and total annual streamflow generation
normalized by respective control simulations (vertical axis) for all year and each snowpack
distribution without precipitation forcing. Solid black line indicates a ratio of 1, with dots
above (below) the line indicating an increase (decrease) in streamflow generation.
The shape of the cumulated streamflow hydrograph Figure 8 is largely similar when
precipitation forcing is withheld relative to Figure 5, however with greater variability—more
© ASCE
World Environmental and Water Resources Congress 2019
222
Downloaded from ascelibrary.org by University of Liverpool on 08/26/19. Copyright ASCE. For personal use only; all rights reserved.
World Environmental and Water Resources Congress 2019
uniform snowpacks produce more early-season streamflow and less uniform snowpack
producing more late-season streamflow. Compared with the 2.2% decline in Figure 5, there is a
decrease of 12.1% in total streamflow generation for the uniform case relative to the control
when excluding precipitation, nearly a 10% difference from the previous set of simulations.
Additionally, the date of peak streamflow occurs 7 days earlier when using the mean snowpack
distribution, an increase of 5 days.
Perhaps most interesting, when precipitation forcing is withheld, snowpack uniformity
always produces less streamflow relative to the control simulation (Figure 9). This is in contrast
to the first scenario (Figure 6) that included precipitation forcing, where deeper snowpacks had
instances of increased uniformity producing more streamflow, suggesting that deeper snowpack
years may have experienced enhanced melt-season precipitation that confounded the role of
snowpack uniformity. With the exclusion of precipitation, there are decreases by as much as
17.2% in total streamflow generation when using the mean snowpack distribution, a reduction
that is 15% greater than the previous simulations. While there does appear to be a slight trend (R 2
of 0.25) towards larger snowpacks showing smaller streamflow decreases in the absence of
precipitation, there is still variability from year-to-year, probably due to temperature differences
from year-to-year.
While trying to discern the physical mechanisms as to why differing snowpack distributions
alter catchment hydrology poses inherent challenges in a modeling framework such as this one,
but several processes can lend themselves to driving these changes. Across all simulations, the
total amount of snowmelt is largely conserved because each set of iterations have the same initial
mean snowpack (i.e. they all have the potential to produce the same streamflow). A consistent
finding is that early-season melt is much more prevalent for increasingly uniform snowpacks. A
greater persistence of snow and later peak streamflow timing was noted for more variable
snowpacks. Coupled with this timing of snowmelt is a 2% increase for June 1 soil moisture for
the mean snowpack distribution and a water table that is closer to the surface. With the increased
storage of water in the soil, as opposed to in the snowpack for the control distributions, there is
an increase in total evapotranspiration of water from the soil – 0.5% increase for simulations
with precipitation, 6% increase for simulations with no precipitation – as the snowpack becomes
more uniform. This eventual loss of water to the atmosphere is a key mechanism leading to
declines in streamflow generation for more uniform snow conditions.
CONCLUSION
This study has highlighted the role of snowpack variability in streamflow generation, while
subsequent summer meteorology can ultimately change the impacts of various snowpack
distributions, there is evidence that more uniform snowpacks will reduce overall catchment water
yield. While more uniform snowpacks tend to produce less streamflow form the catchment,
increased SWE conditions tend to mitigate these reductions.
Kittel et al. (2015) notes that while there is an observed precipitation trend of 60 mm year -1
decade-1 for our study domain, there is also an observed temperature trend of 0.8°C decade -1
occurring simultaneously. While increases in precipitation may alleviate increased potential
evaporation demands due to increasing temperate, the increasing temperature may alter
snowpack wind redistribution. While this region has experienced increases in both precipitation
and temperature, other regions will experience differing changes to their respective climates that
could exacerbate the changes in streamflow generation if just increased temperatures occur. This
altered snowpack distribution appears to make streamflow generation less efficient, thus
© ASCE
World Environmental and Water Resources Congress 2019
223
World Environmental and Water Resources Congress 2019
Downloaded from ascelibrary.org by University of Liverpool on 08/26/19. Copyright ASCE. For personal use only; all rights reserved.
decreasing the amount of water available to the areas dependent on these seasonal snowpacks for
their water resources needs.
It is of note that this study looks at a singular alpine catchment and the results might vary
when investigating other watersheds. There is a need to continue this sort of analysis with GCM
output and remotely sensed snow cover data to compare sensitivities to snowpack distribution
homogeneity across other alpine catchments, as it could provide meaningful information as to
how water resources might be shifting in the future.
ACKNOWLEDGMENTS
This research was supported by funding from the National Science Foundation’s Division of
Environmental Biology (NSF / DEB 1027341).
REFERENCES
Abbaspour, K. C., Yang, J., Maximov, I., Siber, R., Bogner, K., Mieleitner, J., ... & Srinivasan,
R. (2007). Modelling hydrology and water quality in the pre-alpine/alpine Thur watershed
using SWAT. Journal of hydrology, 333(2), 413-430.
Caine, N. (1996). Streamflow patterns in the alpine environment of North Boulder Creek,
Colorado Front Range. ZEITSCHRIFT FUR GEOMORPHOLOGIE SUPPLEMENTBAND,
27-42.
Clow, D. W. (2010). Changes in the timing of snowmelt and streamflow in Colorado: a response
to recent warming. Journal of Climate, 23(9), 2293-2306.
Essery, R., & Pomeroy, J. (2004). Vegetation and topographic control of wind-blown snow
distributions in distributed and aggregated simulations for an Arctic tundra basin. Journal of
Hydrometeorology, 5(5), 735-744.
Field, C. B., Barros, V. R., Mach, K. J., Mastrandrea, M. D., van Aalst, M., Adger, W. N., ... &
Birkmann, J. (2014). Technical summary. In Climate change 2014: Impacts, adaptation, and
vulnerability. Part A: Global and sectoral aspects. Contribution of working group II to the
fifth assessment report of the intergovernmental panel on climate change (pp. 35-94).
Cambridge University Press.
Godsey, S. E., Kirchner, J. W., & Tague, C. L. (2014). Effects of changes in winter snowpacks
on summer low flows: case studies in the Sierra Nevada, California, USA. Hydrological
Processes, 28(19), 5048-5064.
Graham, L. P., Hagemann, S., Jaun, S., & Beniston, M. (2007). On interpreting hydrological
change from regional climate models. Climatic change, 81(1), 97-122.
Gurtz, J., Baltensweiler, A., & Lang, H. (1999). Spatially distributed hydrotope-based modelling
of evapotranspiration and runoff in mountainous basins. Hydrological processes, 13(17),
2751-2768.
Hamlet, A. F., Mote, P. W., Clark, M. P., & Lettenmaier, D. P. (2005). Effects of temperature
and precipitation variability on snowpack trends in the western United States. Journal of
Climate, 18(21), 4545-4561.
Huntington, J. L., & Niswonger, R. G. (2012). Role of surface‐water and groundwater
interactions on projected summertime streamflow in snow dominated regions: An integrated
modeling approach. Water Resources Research, 48(11).
Jepsen, S. M., Molotch, N. P., Williams, M. W., Rittger, K. E., & Sickman, J. O. (2012).
Interannual variability of snowmelt in the Sierra Nevada and Rocky Mountains, United
States: Examples from two alpine watersheds. Water Resources Research, 48(2).
© ASCE
World Environmental and Water Resources Congress 2019
224
Downloaded from ascelibrary.org by University of Liverpool on 08/26/19. Copyright ASCE. For personal use only; all rights reserved.
World Environmental and Water Resources Congress 2019
Jones, J. A., Creed, I. F., Hatcher, K. L., Warren, R. J., Adams, M. B., Benson, M. H., ... &
Clow, D. W. (2012). Ecosystem processes and human influences regulate streamflow
response to climate change at long-term ecological research sites. BioScience, 62(4), 390404.
Kittel, T. G., Williams, M. W., Chowanski, K., Hartman, M., Ackerman, T., Losleben, M., &
Blanken, P. D. (2015). Contrasting long-term alpine and subalpine precipitation trends in a
mid-latitude North American mountain system, Colorado Front Range, USA. Plant Ecology
& Diversity, 8(5-6), 607-624.
Livneh, B., Deems, J. S., Schneider, D., Barsugli, J. J., & Molotch, N. P. (2014). Filling in the
gaps: Inferring spatially distributed precipitation from gauge observations over complex
terrain. Water Resources Research, 50(11), 8589-8610.
Livneh, B., Deems, J. S., Buma, B., Barsugli, J. J., Schneider, D., Molotch, N. P., ... & Wessman,
C. A. (2015). Catchment response to bark beetle outbreak and dust-on-snow in the Colorado
Rocky Mountains. Journal of Hydrology, 523, 196-210.
McGuire, C. R., Nufio, C. R., Bowers, M. D., & Guralnick, R. P. (2012). Elevation-dependent
temperature trends in the Rocky Mountain Front Range: changes over a 56-and 20-year
record. PLoS One, 7(9), e44370.
Mote, P. W., Hamlet, A. F., Clark, M. P., & Lettenmaier, D. P. (2005). Declining mountain
snowpack in western North America. Bulletin of the American meteorological Society, 86(1),
39-50.
Nash, J. E., & Sutcliffe, J. V. (1970). River flow forecasting through conceptual models part I—
A discussion of principles. Journal of hydrology, 10(3), 282-290.
Pierce, D. W., Barnett, T. P., Hidalgo, H. G., Das, T., Bonfils, C., Santer, B. D., ... & Wood, A.
W. (2008). Attribution of declining western US snowpack to human effects. Journal of
Climate, 21(23), 6425-6444
Scherrer, D., & Körner, C. (2011). Topographically controlled thermal‐habitat differentiation
buffers alpine plant diversity against climate warming. Journal of biogeography, 38(2), 406416.
Stewart, I. T., Cayan, D. R., & Dettinger, M. D. (2005). Changes toward earlier streamflow
timing across western North America. Journal of climate, 18(8), 1136-1155.
Suding, K. N., Farrer, E. C., King, A. J., Kueppers, L., & Spasojevic, M. J. (2015). Vegetation
change at high elevation: scale dependence and interactive effects on Niwot Ridge. Plant
Ecology & Diversity, 8(5-6), 713-725.
Vaughan, D. G., Comiso, J. C., Allison, I., Carrasco, J., Kaser, G., Kwok, R., ... & Rignot, E.
(2013). Observations: cryosphere. Climate change, 2103, 317-382.
Wigmosta, M. S., Vail, L. W., & Lettenmaier, D. P. (1994). A distributed hydrology‐vegetation
model for complex terrain. Water resources research, 30(6), 1665-1679.
Williams, M. W., Losleben, M., Caine, N., & Greenland, D. (1996). Changes in climate and
hydrochemical responses in a high‐elevation catchment in the Rocky Mountains, USA.
Limnology and Oceanography, 41(5), 939-946.
© ASCE
World Environmental and Water Resources Congress 2019
225
Документ
Категория
Без категории
Просмотров
0
Размер файла
864 Кб
Теги
022, 9780784482346
1/--страниц
Пожаловаться на содержимое документа