Surface & subsurface processes in mountain environments evapotranspiration snowmelt precipitation infiltration Roger Bales Martha Conklin Robert Rice Fengjing Liu Peter Kirchner runoff sublimation ground & surface water exchange Mountain Hydrology Research Group, Sierra Nevada Research Institute, UC Merced
Understanding hydrologic processes in seasonally snow-covered mountain basins: our assumptions 1. The basis for process understanding is new measurements 2. Processes are coupled & best studied together 3. Spatial measurement design at the basin scale is a research area 4. Multi-scale measurements are required for scaling process understanding 5. Following snowmelt will yield process insight: ET snow distribution melt timing infiltration partitioning recharge runoff
Topics covered in this talk Snowmelt generation & sources Identifying streamflow sources & contributions Stream & soil moisture response to ET Measurement design in mountain basins
Snow distribution & snowmelt generation 3 snow pillows in Merced R. basin, 7 in Tuolumne Insufficient for spatial snow estimation
Operational snow measurements: not representative 90 Gin Flat data 90 80 snow course mean 80 SWE, cm 70 SWE, cm 70 60 1-km 2 spatial mean 60 50 50 February 2006 April 1, 2006 Snow pillow & snow course measurements provide limited information on spatial distribution patterns. Sites are not representative of the terrain & thus fail to represent basin-wide snow depth or water equivalent.
Plot-scale controls on snow distribution 3300 3390 3450 Intense surveys provide guidance for placement of automatic instruments 112 106 100 94 N 88 82 elevation solar radiation slope 2 km 4 km wind exposure vegetation density snow depth, cm relations differ at small catchment vs. regional watershed scale 76 70 64 58 Molotch et al., 2004
Snow covered area (SCA) from MODIS % SCA 76-100 51-75 26-50 1-25 Tuolumne Merced DOY 131 May 10, 2004
Estimating snow water equivalent (SWE) & snowmelt from SCA time series Determine daily potential snowmelt per elevation band based on temperature index (or energy balance) calculation Potential snowmelt quantity applied to all areas with snowcover, using fraction SCA per pixel That is, if an area has snowcover it is assumed to contribute melt at a rate equal to the potential snowmelt times SCA Amount of snowmelt calculated up to day when SCA is depleted equals SWE accumulation
Melt season SCA 2004: snow 85% of average ground became snow free relatively early Each higher elevation band requires ~1 mo longer to become snow free 2005: snow 165% of average ground became snow free ~1 mo later than in 2004 SCA fraction SCA, fraction SCA fraction 1.0 0.8 0.6 0.4 0.2 Tuolumne 2004 elev band, m 3750 3450 3150 2850 2550 2250 1950 1650 0.0 3/1/04 4/1/04 5/1/04 6/1/04 7/1/04 8/1/04 9/1/04 10/1/04 Merced 2005
60 SWE based on SCA plus degree day snowmelt Volumne snowmelt, 10 7 m 3 50 40 30 20 Based on SCA depletion and degree day calculation Tuolumne 2004 3450 3150 2850 2550 10 2250 1950 Cumulative melt also calculated based on interpolated SWE More rapid depletion Interpolation over estimates below 3,000 m & under-estimates above 3,000 m Volumne snowmelt, 10 7 m 3 0 3/1/04 4/1/04 5/1/04 6/1/04 7/1/04 60 50 40 30 20 10 Based on interpolated SWE 3450 3150 2850 2550 2250 1950 0 3/1/04 4/1/04 5/1/04 6/1/04 7/1/04
SWE based on SCA plus degree day snowmelt Merced, 2005 Interpolated SWE Shows rapid depletion at all elevations Less total melt for interpolated method because snow pillow sites melted out early
Fraction of snowmelt from various elevations Merced basin similar in 2005 Similar pattern in Tuolumne
What paths does snowmelt take in a basin? overland flow subsurface flow groundwater
Compare 2006 (wet) & 2007 (dry) on Merced R. Head of Yosemite Valley: snow dominated Below Yosemite in Merced R. Canyon: some rain Snowfall
Sampling locations ions & isotopes
Relative end-member contributions from 3-component mixing model spring summer & fall winter overland flow lateral subsurface groundwater snow & surface water soil fractured rock Increasing importance Subsurface flow & groundwater similar in 2006 & 2007
Wolverton Creek, Sequoia NP: stream stage Base - flow August 14 - September 7, 2007 Base-flow period, 2150 m elevation stage cm 25 20 15 10 5 15 20 25 30 5 August September 0
Wolverton Creek, Sequoia NP: stage & precipitation Base - flow August 14 - September 7, 2007 Base-flow period, 2150 m elevation stage cm precip.001mm 25 20 15 10 As we add measurements and integrate them causal relationships become increasingly clear. 15 20 25 30 5 August September 5 0
Wolverton Creek, Sequoia NP: stage, precip & air temperature Base - flow August 14 - September 7, 2007 Base-flow period, 2150 m elevation stage cm air temp C precip.001mm 25 20 15 3 to 6 hour lag between daily low flow and Tmax or daily high flow and Tmin. 10 5 15 20 25 30 5 August September 0
Wolverton Creek, Sequoia NP: stage, precip, air temp & sap flow Base - flow August 14 - September 7, 2007 Base-flow period, 2150 m elevation stage cm sap-flow C air temp C precip.001mm 25 20 Peak sapflow occurs at solar noon about 1 hr before Tmax. 15 10 5 15 20 25 30 5 August September 0
Soil volumetric water content response to snowmelt: Wolverton basin Soil Moisture Response 0.25 0.2 0.15 0.1 0.05 0 drip edge under canopy 160 140 120 100 volumetric fraction cm snow SWE snow depth 80 60 40 20 Soil volumetric water content Snow & SWE depth, cm 0 3/15/2007 14:00 3/20/2007 14:00 3/25/2007 14:00 3/30/2007 14:00 4/4/2007 14:00 4/9/2007 14:00 4/14/2007 14:00 4/19/2007 14:00 4/24/2007 14:00 4/29/2007 14:00 5/4/2007 14:00 5/9/2007 14:00 5/14/2007 14:00 5/19/2007 14:00 5/24/2007 14:00 March April May 2007 60cm DE 60cm UC Snow Depth cm SWE Snow depth from acoustic sensors over each pit Snow density from Panther Meadow snow course
Designing better ground-based measurement networks for process research & applications Spatial measurement strategy: strategically placed ground-based instrument clusters, plus satellite data
Sierra Nevada Hydrologic Observatory Tahoe NF Yosemite NP Sierra NF Sequoia NP http://snri.ucmerced.edu A new generation of measurements & infrastructure Collaboration: universities, state & federal agencies, water districts Initial instrument clusters near rain-snow transition Measurement strategies rely heavily on satellite remote sensing coupled with new in situ measurements
Prototype instrument cluster: Wolverton, Sequoia National Park
10 km2 Wolverton basin stream gage water balance instrumentation met station meadow transects
Stream stage & discharge Stream pressure transducers installed in summer 2006 Rating curves in progress Adjacent, higher elevation (Tokapah) basins already monitored
Wolverton meteorological stations Two stations installed in fall 2006, at Wolverton & Panther Meadow
Wolverton, Sequoia NP Vertical soil profiles of moisture & temperature, colocated with snow & radiation Random vs. stratified placement relative to canopy Variables: N/S, elevation, vegetation Wells & piezometers in meadows & riparian areas Two stations installed in fall 2006, at Wolverton & Panther Meadow
Observatory as transect: Tioga Road corridor, Yosemite National Park Instrument sites Instrument sites leverage operational & research investments 2700 2400 Strategy: rather than spreading 2100 instruments across a whole basin, this transect statistically samples the 1800 variability in the Tuolumne & Merced basins, taking advantage of the Tioga 1500 Pass Road as infrastructure 0 10 20 30 40 50 60 70 80 90 kilometers
Gin Flat site, Tioga Pass road, Yosemite NP Wireless pod for recording snow depth Capturing variability across the landscape Instruments placed in km-scale clusters along elevation & latitude transects, linked by wireless pods to a mother pod with telemetry capability Add new measurements at operational sites
Embedded sensor network for mountain water cycle signal/data to/from other nodes One node Sensors snow depth air temperature relative humidity solar radiation soil moisture soil temperature Pod microcomputer data storage radio battery solar cell signal/data Mother pod microcomputer radio battery solar cell data logger & IP connection via phone, radio or direct network data & control Pod from Sensornet, Inc. signal/data to/from data system
Final thoughts on measurements & process understanding Snowmelt generation & sources: strategically place instrument clusters to blend w/ satellite snowcover Identifying streamflow sources & contributions: highfrequency geochemical sampling over representative time periods Stream & soil moisture response to ET: high-frequency data from multi-parameter, strategically placed instrument clusters Measurement design in mountain basins: gradients in elevation & latitude, at multiple scales
Remote sensing work supported by NASA Collaborators: J. Dozier (UCSB) & T. Painter (U. Utah) UCM team: X. Meng, M. Meadows Instrument clusters supported by NSF, DWR & UC Process research supported by LLNL, CEC, NSF Hydrologic observatory collaborators: T. Harmon, Q. Guo UCM J. Kirchner, B. Boyer UCB J. Hopmans, G. Fogg UCD C. Tague UCSB N. Molotch UCLA M. Goulden UCI D. Cayan, M. Dettinger UCSD D. Johnson UNR USFS-PSW, SNP, YNP, SNF