2012 AGU Chapman Conference on Remote Sensing of the ...

has more improved resolution ( ) to compare with LDASresolution ( ) and produces optimized observation data,satellite data, and model’s results. This dataset is anattractive alternative to estimate for constrained land fluxesand variables due to strong heterogeneous land. Throughthese processes, the CLM calculate hydro-meteorologicalvariables such as net radiation, latent, sensible, and groundheat fluxes. For this study, the comparison of observationdata and models’ results should be verified for model’sapplicability in the East Asia.Kim, EdwardMultilayer Snow Microwave ModelIntercomparisons and Scale Implications for FutureSnow MissionsKim, Edward 1 ; DURAND, Michael 2 ; MARGULIS, Steven 3 ;MOLOTCH, Noah 4 ; Lemmetyinen, Juha 5 ; Picard, Ghislain 61. NASA GSFC, Greenbelt, MD, USA2. Ohio State University, Columbus, OH, USA3. UCLA, Los Angeles, CA, USA4. Univ. of Colorado, Boulder, CO, USA5. Finnish Meteorological Institute, Helsinki, Finland6. LGGE, Grenoble, FranceAbstract: Microwaves are well-suited to the task of snowremote sensing with their high sensitivity to snow extent andsnow water equivalent, and particularly since, unlike visiblesensors, microwave sensors do not require solar illuminationand can see through cloud cover. But the same highsensitivity also creates more stringent requirements onknowledge of snow pack characteristics or other constraints,as the signal is a path-integrated quantity and so it ispossible for multiple conditions to produce the samesignature. e sensitivity has been investigated by numerousresearchers, and the corresponding requirements onknowledge of snow characteristics is the topic of relatedabstracts. Snow is also a complex structure from a microwaveradiative transfer point of view. The natural episodic arrivalof snow, snow grain metamorphism, and melt/refreeze cyclescan create layering and other microstructural variations thatare easily seen by microwave sensors. Understanding thesesignatures usually requires a radiative transfer models withmultiple layers and within each layer a means of representingthe size and/or distribution of snow grains or correlationlength. Over the past few years, several groups have collectedin situ snow and microwave brightness temperaturemeasurements to aid in the improvement of microwaveradiative transfer models of snow. In this paper, fieldmeasurements from the US, Canada, and Finland will beused to drive a number of multilayer snow microwaveradiative transfer models, including MEMLS, multilayerHUT, and others. The outputs will be compared to exploreeach model’s response to the different snow conditions.Most importantly, we are interested in ascertaining themodel complexity required to achieve a given accuracy (e..g, 5K), and the corresponding requirements on accuracy of theinput parameters. For retrievals based on model inversion ordata assimilation approaches, this is a key question. Theanswers directly impact the sensitivity and accuracyrequirements of sensors on future snow missions. Intimatelyconnected to this is the spatial resolution of theobservations and retrievals. For the same snow conditionsand same radiative transfer model or retrieval scheme, therequirements on sensors and knowledge of snow conditionscan vary significantly as a function of spatial resolution.This is perhaps more so for snow than for other hydrologicalretrievals, such as for soil moisture. We will present theresults of scaling tests, using the above models and snowmeasurements to help understand the implications forfuture snow missions.Kirchner, Peter B.Measuring under-canopy snow accumulation withairborne scanning LiDAR altimetry and in-situinstrumental measurements, southern SierraNevada, CaliforniaKirchner, Peter B. 1 ; Bales, Roger C. 1 ; Molotch, Noah P. 21. Sierra Nevada Research Institute, University of California,Merced, Merced, CA, USA2. Institute of Arctic and Alpine Research, University ofColorado at Boulder, Boulder, CO, USASnow distribution Estimates in forested environmentsdemonstrate a high level of uncertainty due to the inabilityof remote sensing platforms to observe reflectance underdense vegetation and the limited availability of spatial andtemporal in-situ measurements. Thus measuring snowunder forest canopies remains an unresolved problem forremote sensing of snow cover in forested landscapes. In thisstudy we carefully analyzed filtered paired snow on and snowoff scanning LiDAR altimetry collected in the 2010 wateryear, from the Kaweah River watershed, Sierra Nevada,California, to establish snow depths over a 52.5 squarekilometer area covering a wide range of slopes aspects,elevations and forest types including Giant Sequoia groves,mixed conifer and sub alpine forests. Using 1 m 2 meanelevation grids produced from filtered first and last returnswe established a distinction between snow in open areas andthose under the canopies by selecting areas where meanground and canopy returns overlapped defining the canopyedges of mature trees and the under canopy of small treesand shrubs. In addition we analyzed in-situ time series dataof snow depth density precipitation, temperature, andupstream bright band radar data to establish a deeperprocess understanding of the dynamics between snowaccumulation in the open and under forest canopies. Resultsindicate a decrease in under canopy depth at all locations,but lower elevations demonstrate a greater decrease and a10% higher coefficient of variation in snow depth and an 8%increase in density. Upstream bright band radar and metdata from hydrologic observatory sites indicate the locationsof increased variability in depth and higher density receiveda greater percentage of precipitation as rain. Our findingsprovide a metric for estimating under canopy snowaccumulation where it cannot be directly observed directlywith remote sensing and suggest the elevation of the rain83