Topoclimatic Influences on Climatic Water Balance in Complex Terrain: Implications for modeling Species Occurrence, Regeneration and Productivity Zachary A. Holden, Solomon Dobrowski and John Abatzoglou Scaling Climate in Mountainous Terrain • Mountains create steep biophysical gradients • Radiation • Temperature lapse rates • atmospheric humidity • snowmelt timing • Wind speed • These factors will influence: • • • • Seed germination Regeneration Species redistributions Site-specific Leaf area and productivity Vegetation in Complex Terrain Solar Radiation is King • Topographic shading/variable sun angles with • Time of year influence air temp, soil temperature, snowmelt, humidity…. Scale and Resolution GCM/RCM data typically downscaled to 10-15 km resolution PRISM data produced at 800meters – 4km Many forest management decision require data at the resolution of terrain - Seed stock selection; growth; shifts in species occurrence 6-8 deg. C variation Within a 1 km2 window In complex terrain Correlation Matrix 10 RAWS Bitterroot NF TMAX Strong Regional Coherence in Daytime temperatures Solar heating during the day leads to well mixed air decreasing temperatures with elevation (lapse rates) Maximum Daily Temperature at 10 Bitterroot NF RAWS Nighttime Decoupling from Free Atmosphere Correlation Matrix 10 RAWS stations on The Bitterroot NF Circled Stations are At mid-slope or valley bottom locations Minimum RH Maximum RH August 12th-21th: Daytime Stable High Pressure system sets up inversion Nighttime lapse rates in Big Creek, Bitterroot NF Average Difference Between Valley bottom sensor and sensors 200 meters above it is 5 Degrees C. 400-500 meter deep mid-slope thermal bands Some nights show full inversions: Temperatures warmer at 8000 feet than at 3500 feet in the valley bottom Bitterroot Valley, August 7th, 2009 Prevailing wind from the west Wind Speed Variation in Complex terrain Wind speed and direction varies with topographic exposure relative to prevailing wind direction Downscaling topographic variation in wind speed • • • • Windninja model (Jason Forthofer; Firelab) Daily wind simulations (June-October; 1979-2010) ~ 700,000 daily simulations averaged to monthly avg. Higher wind speeds on ridges/west-facing mountain fronts; lower in valleys Toward Improved understanding of climate-vegetation dynamics: Development of high spatial resolution climatic water balance models Penman-Montieth equation for evapotranspiration Integrates climate and energy into mechanistic variables Temperature Radiation ET = Atmospheric Vapor Pressure (RH) ∆ 𝑅𝑛 −G +𝜌𝑎 𝑐𝑝 𝑒𝑠 −𝑒𝑎 /𝑟𝑎 𝑟 ∆+𝛾 1+ 𝑠 𝑟𝑎 Aerodynamic resistance (Wind) Each aspect of the Penman-Monteith model varies with terrain The climatic water balance • PET = potential for a site to evaporate water • AET = actual evaporation, given moisture. (Think productivity) • Deficit= The difference. • PET – AET = deficit. • Unmet atmospheric demand Figure from Nathan Stephenson, USGS, AGU presentation 2011 Water Balance Inputs (60m) 30 yr Monthly Average: Min. Temperature Max. Temperature Solar Radiation Precipitation Wind Speed (windninja) Soil water capacity 30 year (1971-2000) Deficit Using high resolution solar radiation grids, we can capture aspect-scale variation in water balance Modeling the influence of wind on evapotranspiration • Selkirk mountains Topographic complexity and implications for influences of wind on climatic water balance Modeling species occurrence using a water balance approach • Presence/Absence data from FIA • GLM: PIAL ~ AET + DEFICIT + TMIN • Limiting factors (energy, water, temperature) PIAL niche space in 5 genetic seed zones defined by Mahalovich Differentiation in where species occur on the spectrum of evapotranspiration and deficit by zone Comparison of Adult vs. 0-1 inch (regenerating) size classes Evidence for shifting The regeneration niche of Whitebark? Regeneration shift toward higher energy (AET) and higher deficits? Some places are seeing DOWNHILL shifts in regeneration?! Interesting….Mechanisms? Visualizing Species niches on the landscape PIEN probability PIPO probability Fine-scale variation with aspect PIEN PIPO Putting Whitebark in the context of other species Niches Linking Water Balance to Productivity Discrete Return LiDAR Clear Creek Drainage, Nez Perce NF 80,000 acres LiDAR metrics at 20m resolution Modeling stand level forest metrics with LiDAR • 62 Fixed radius plots • Empirical models fit • using LiDAR height metrics as as independent variables Clear Creek Drainage Nez Perce NF Basal Area and Crown Bulk Density modeled from LiDAR data Clear patterns of productivity relative to summer deficit Low elevation south slopes limited by water High elevation north slopes limited by energy A topoclimatic monitoring program for the Northwest US • Many partners in the USFS, USGS, and area universities • 2000 sensors distributed from 2009-2011. • 400 RH sensors added in 2010 What can we learn from inexpensive sensors? Thermoworks Logtag ($22) Thermochron ibutton ($35) • How much does topography influence surface air temperatures? • Space-time models of surface air temperature variation that account for variable lapse rates, cold air drainage and aspect A gill-style radiation shield for measuring air temperatures in forested environments $2.00 material costs 12 minute construction time Minimal bias under Forest canopy August 2010 Boise Basin Daily Tmax There is huge variation In lapse rates from day To day. Also dynamic adjust. In how much north vs. south Slopes vary depending on Humidity and sky conditions PRISM derived water balance vs. with microsensors Valley bottoms much wetter when we account for CAD Wind-exposed slopes much drier Discussion • Climatic water balance provides a physical basis for addressing climate-terrain interactions and relationships with vegetation • Allows us to address climate-vegetation relationships in a mechanistic way, while still accounting for topography • Obviates the need for using gradient models and empirical indices Future work in progress: • Use distributed sensor networks to downscale General Circulation Model predictions for temperature and relative humidity • Develop species occurrence and regeneration models under a range of future climate scenarios • Integrate climatic water balance models with LiDAR data to develop LiDAR-based structure models applicable to a range of forest types Next Steps…integrating high resolution microclimate data into water balance Predicted Minimum temperature, Big Creek, July 24th