the effect of vertical canopy structure on snow processes · •unperturbed forest versus •open...

The Effect of Vertical Canopy Structure on Snow

Processes

Laura E. McGowan,

Kyaw Tha Paw U, Helen Dahlke, & William Massman

Forested snow cover

• Critical driver of energy & water budgets

• A natural reservoir, providing Western US with agricultural water, drinking water, & hydropower

• Yet accurate estimates of snow cover & snowmelt in forested areas remains a challenge

Below canopy snowpack

Precipitation

Soil

Snow-forest processes complicate modeling


Precipitation

Throughfall

Soil



Precipitation

Throughfall

Interception

Soil



Precipitation

Throughfall

Interception

Sublimation/Evaporation

Soil



Precipitation

Throughfall

Interception


Soil


Unloading and Drip


Precipitation

Throughfall

Interception


Evaporation/Sublimation

Soil


Unloading and Drip


Precipitation

Throughfall

Interception

Melt


Evaporation/Sublimation

Soil


Unloading and Drip

Vertical structure influences snow processes & modeling

• Controls snow’s spatiotemporal distribution

• Alters energy & water budget

• Resolving vertical canopy structure critical to understanding snow dynamics!

• However few studies explicitly model complex processes in multiple vertical layers!

Most studies only look at planar variables

Methods to examine snow canopy processes

• Unperturbed forest versus • Open area or lake1

• Clear cut forests2

• Post wildfire3

• Mountain pine beetle damage4

• Variation in canopy parameters• Leaf area indices5

• Canopy closure6

Multilayer model performance

PART I

• Accurately reproduce snow water & energy budget?

• Outperform a standard single-layer model ?

PART II

• Quantify impact of varied vertical structure on water & energy budgets

Two sets of simulations

1. Multilayer simulation

2. Single-layer simulation

Temperature, humidity, shortwave radiation, CO2, wind speed, pressure, precipitation

10 layerswithin-canopy

4 soil layers

10 layersabove-canopy

Image modified from Kent 2015

9 sunlit leaf angles1 shaded leaf angle

Energy budget, temperature, physiology, radiative transfer

equations & water budget

Snow

Temperature, humidity, shortwave radiation, CO2, wind speed, pressure, precipitation

1-layer model

4 soil layers

10 layersabove-canopy

Image modified from Kent 2015

9 sunlit leaf angles1 shaded leaf angle

Energy budget, temperature, physiology, radiative transfer

equations & water budget

Snow


Modeled snow processesPrecipitation

ThroughfallInterception

Melt

Unloading & drip

Evaporation/sublimation

Evaporation/sublimation

Soil

- -- Measured - - -Multilayer Single-Layer

Multilayer better agreement with measurements

• Snow depth had better agreement with measurements magnitudes, forecast errors, and correlations• Also the case for the energy fluxes

• Canopy-top net radiation simulated by the multilayer model were in closer agreement to measurements

• As well as turbulent flux components

Taylor 2001 plot of the normalized standard deviation,centered RMSE, and correlation coefficient

Multilayer vs single-layer canopy model

• Multilayer accurately captures snow water & energy budgets on daily & annual scales

• Multilayer reduced forecast error & greater correlation to the measurements for both snow depth & energy components

Quantify impact of varied vertical structure on water & energy budgets

Reduced snowpack & shorter snow season for top-heavy canopies?

• More snow held towards the top of the canopy

• Increased winds at the top of canopy

7 vertical canopy architectures

3 controls

1) Bottom-Heavy

2) Top-Heavy

3) Even

4 old-growth coniferous trees*

4) Tree ‘98’

5) Tree ‘1137’

6) Tree ‘Minerva’

7) Douglas fir composite

*Massman 1982

7 vertical canopy architectures:All other canopy parameters identical

Canopy parameters

7 vertical canopy architectures:All other canopy parameters identical

Canopy parameters

• 1-sided total plant area indices

• Canopy architecture

• Leaf/canopy drag coefficient

• Canopy height

• Near infrared reflectivity

• Visible leaf reflectivity

• Aerodynamic drag coefficient

Bottom-heavy= 6.0 Top-heavy= 6.0

Differing vertical structures had significantly different annual snowpack water availability

• Max snow depths varied directly with vertical biomass density distribution

• Canopies with greatest biomass in the top half of the canopy had the greatest reduction in snow depth & earliest melt-out • ‘Minerva’ had 40% reduction in snow depth

thickness (0.61 m) and a 20 day earlier melt out compared to the predominately bottom-heavy canopy (control Bottom-heavy)

• Top-heavy architecture led to conditions favoring increased evaporative & sublimation losses • Snow within the canopy & beneath the

canopy

Mean

Date Δ

ACASA Bottom-Heavy 30-Jun

Even Distribution 22-Jun 7.4

Tree '1137' 19-Jun 11.1

Tree '98' 18-Jun 11.3

ACASA Top-Heavy 15-Jun 14.5

Composite Tree 16-Jun 13.6

Tree 'Minerva' 9-Jun 20.4

Increased latent heat loss in top-heavy canopies

Top-heavy canopies have increased loss of water to the

atmosphere

• ~40% decrease in peak snowpack depth (0.6 m)

• ~2.5 weeks reduction in snow season length

• Therefore, accounting for vertical canopy structure & processes may be a critical step for improving estimates of annual water budgets from forested areas

Thank you!Any Questions?

Acknowledgement to NSF award EF1137306/MIT subaward 5710003122 to the University of California, Davis

References

1. Hardy et al., 1997; Mahat & Tarboton, 2014; Broxton et al., 2014; Harpold et al., 2014; Harding & Pomeroy, 1996

2. Schmidt & Troendle, 1989; Golding & Swanson, 1986; Jost, et al., 2009; Berndt, 1965; Berris & Harr, 1987; Anderson & Gleason

3. Teti, 2008; Seibert et al., 20104. Bewley et al., 2010; Boon, 2007, Teti, 2008; Pugh & Small, 2012; Pugh & Gordon, 20125. Luo et al., 2008; Pomeroy et al., 2002; Davis et al., 19976. Ellis et al., 2013; Essery, 1998; Musselman et al., 2012; R. Winkler & Moore, 20067. Massman, 19828. Taylor, 2001

the effect of vertical canopy structure on snow processes · •unperturbed forest versus •open...

Documents