H31E-0693
Large-Eddy Simulations of Forest Canopy Micro-Scale Structural Heterogeneity Effects on the Atmospheric Boundary Layer
The recently developed Regional Atmospheric Modeling System (RAMS)-based Forest Large-Eddy Simulation (RAFLES) is used to explore the effects of micro-scale structural heterogeneity of forest canopies on the Atmospheric Boundary Layer. This heterogeneity is represented by structural features such as differences between the heights and leaf densities of individual trees. The autocorrelation length scale of these features is smaller than the mean canopy height. The virtual canopy generator (V-CaGe) is used to generate a realistic canopy, based on observed canopy statistics from a hardwood stand in the Duke Forest, as a control case. Two different environments are considered: one is in the winter season characterized by moderate winds aloft, a sparse canopy and a slightly unstable boundary layer; the other is in the spring season with a dense canopy and a strongly convective boundary layer. The control case (under the two environments) is compared with two test cases simulating different levels of heterogeneity, with the same mean canopy properties and atmospheric forcing: (1) A horizontally homogeneous canopy; and (2) A heterogeneous canopy that also includes tree fall gaps. Micro-scale heterogeneity impacts the roughness properties, the effective drag coefficients, the displacement height, and the planar-averaged flow statistics. It also leads to higher-order effects pertaining to the spatial statistics of the ejection-sweep cycle generating correlations between micro-scale canopy features, flow statistics and fluxes, which persist up to five times the height of the canopy.
H31E-0694
Influence of surface heterogeneity on a realistic convective boundary layer
Turbulent interactions of momentum and scalars within the convective boundary layer (CBL) over a heterogeneous surface have strong influence on both local flow properties and the regional weather and climate. Based on results of simulations over artificially idealized surface conditions (surface temperature, fluxes, roughness height, etc.), many studies have been performed to build our understanding of influence of surface heterogeneity by analyzing some statistics of characteristic properties in the atmosphere boundary layer. This research attempts to further our understanding by using realistic heterogeneous surface flux fields obtained from the Soil-Moisture Atmosphere Coupling Experiment 2002 (SMACEX) as boundary conditions in a Large-Eddy Simulation (LES). A set of realizations exactly replicating the mean and covariance of surface characteristics (relative to the actual fields at SMACEX) are used in the LES as boundary conditions. Specifically, two parts of analysis are discussed in this study: 1) the characteristic statistics of CBL properties (momentum, scalars, fluxes, variance, etc.); and 2) vertical scales of the CBL (momentum/scalar blending heights, internal boundary layer, convective scaling, etc.). Results from this study not only illustrate the influence of surface heterogeneity on a realistic CBL structure and statistics, but also offer a possibility to develop better parameterizations of surface characteristics used in regional climate models in the future.
H31E-0695
Observations of Evaporation with a New-Generation Raman Lidar
A new atmospheric boundary layer Raman lidar has been designed and built at EPFL. Temperature and humidity can be measured out to 500 m at 1 m spatial resolution, 1 s temporal resolution. A unique multi-telescope design along with 3 nanoseconds 100 Hz pulsed laser operating at 266 nm (in the solar blind region) with ozone correction allows for nearly constant signal to noise, daytime operation and measurement scales of interest in micrometeorology. In this presentation we will detail the underlying principle of this lidar and present new observations of water and temperature microstructure of the lower atmosphere over complex terrain. First calibration results, field observations over a vineyard and time series of vertical temperature and humidity profiles will be presented. http://eflum.epfl.ch
H31E-0696
Study of vegetation effects on land-atmospheric interaction using the Weather Research and Forecasting model and MODIS data
Land-atmospheric interactions have been an important issue in weather and climate modeling. Evapotranspiration (ET) is a major feedback from the land surface to the atmosphere and we analyze land surface processes simulated from the Noah (National Centers for Environmental Prediction – Oregon State University – Air Force - National Weather Service Office of Hydrology) Land Surface Model (Noah LSM) coupled with a generic mesoscale model in the Weather Research and Forecasting (WRF) framework. The International H2O project (IHOP) area is the study domain chosen due to the availability of ground observations. Soil moisture and skin surface temperature (TSK) initializations are tested for the model improvement through the comparison to the ground observation data from the 9 Integrated Surface Flux Facilities (ISFF) over the IHOP area. Three days are selected during the study period - DRY1, WET, and DRY2. The relationship between TSK and Normalized Difference Vegetation Indices (NDVI) is assessed to investigate biophysical effects on ET and the model sensitivity to different model initialization. This study indicates that the soil moisture initialization improves soil moisture simulations but causes overestimations of ET in vegetated areas. We further test the parameterization of vegetation fraction derived from daily or semi-monthly Moderate Resolution Imaging Spectroradiometer (MODIS) NDVI to improve the model sensitivity to soil moisture variations. In addition, we evaluate the applicability of vegetation water content which is not involved in the surface flux algorithm but substantially influential on ET.
H31E-0697
Soil Moisture Patterns and the Lower Atmosphere During the LUCE Measurement Campaign, Switzerland
The LUCE (Lausanne Urban Canopy Experiment) is a measurement campaign which took place on the EPFL campus. This experiment aimed at better understanding micrometeorology in an urban environment. It requires high temporal and spatial density measures in order to cover the heterogeneous areas. The measurement system was based on a wireless sensor network of 90 sensorscope sensing stations which were deployed on the EPFL campus and measured key environmental quantities at high spatial and temporal resolution. This innovative system was deployed starting from November 2006 until June 2007. In addition a SODAR/RASS, a scintillometer and two sonic anemometers coupled with a hygrometer were installed over the campus. Each sensorscope sensing station use the Echo Probe EC-5 sensor to measure soil moisture and the Watermark 200SS Soil Water potential sensor to measure soil water pressure. At first, the methodology employed to treat the soil measurements consists on a spatial interpolation with ArcGis to see the distribution of soil moisture and soil water pressure over the EPFL campus. Then, soil water characteristic curves for each station have been drawn and temporal stability theory has been adapted to our data set. As result, it was possible to connect the spatial location and the soil moisture content with the soil characteristics. Moreover, we have been able to catch the representative points of the mean soil moisture values and to show the temporarily time-stable sites. Finally, the results were compared with laboratory soil analysis and validated. http://eflum.epfl.ch
H31E-0698
The Role of Surface-Exposed Fractures in Land-Atmosphere Dynamics
Throughout the past two decades, most studies that explored flow and transport processes through surface- exposed fractures, focused merely on the role of these fractures as fast conduits for water, salts and contaminants during intensive rain events, flooding or leakage from contamination sources. Conventional wisdom has assumed that as long as fractures are dry, their role in the hydrological cycle is negligible. This study, however, explores the processes occurring within surface-exposed fractures during the dry season, and shows that their role in hydrological and atmospheric cycles is not negligible. As a result of a study that incorporated theoretical work, laboratory experiments and in situ field measurements, we show the following new mechanism: (1) Cold nighttime temperatures, common to arid environments, create unstable air stratification within fractures where atmospheric air becomes denser than fracture air; and (2) This unstable condition results in convective venting of air between the fracture and the atmosphere. Convective venting leads to the following phenomena: greater exchange of atmospheric gases; higher nighttime evaporation rates vs. daytime; lateral transport of solutes toward fracture surfaces and precipitation as salt crusts; and higher heat loss from the vadose zone. Conditions necessary to trigger convection are more prevalent during the winter. Salts that accumulate on the fracture surface may be transported downward during infiltration events and thus contribute to aquifer contamination. This bypass mechanism is especially important in low permeability zones. Convective venting may also impact the life span of hydraulically active fractures by the accumulation over time of low solubility salts. The mechanism of convective venting of air-filled surface exposed fractures is relevant and important to various aspects of mass and energy transfer between the atmosphere and the vadose zone.
H31E-0699
Modeling of Radiation Transport in the Plant Canopy Reflects the Change in Phenology
Solar radiation in a plant canopy plays an important role in the energy balance on both the plant leaves and soil surface, which controls the water and heat transfer in the soil-plant-atmosphere continuum. A simple model was derived from the two-stream model to estimate two important factors characterizing radiation in a plant canopy, namely, transmissivity of the canopy (τ) and the albedo of the canopy (ref), from the absorption coefficient of leaves (α) and leaf inclination factor (F). To clarify the seasonal variation in α and F with plant growth, season-long observations were conducted in rice fields during three different cropping seasons. Values of α were almost constant throughout the growing period; however, values of F tended to increase with change in phenology. Values of F were larger than 0.5 (the theoretical value for random leaf distribution) in the late growth stage due to the alterations in leaf geometry with the change in leaf inclination angle along a more horizontal axis after flowering, while in contrast, values of F were less than 0.5 in the early growth stage due to the distribution bias of leaves. Seasonal variation in F during different cropping seasons could commonly be expressed as a function of developmental stage (DVS). Using this function, τ and ref could be estimated with more accuracy. The proposed radiation model and function is expected to be applicable in more accurate evaluation of the water vapor and heat transfer in the soil-plant-atmosphere continuum that reflects the change in plant phenology.
H31E-0700
Preciptation Parameterization for Idealized Land-Atmosphere Sahel Models
The lead author has previously devised a simple yet powerful numerical model of land-ocean-atmosphere dynamic interactions near a coastline (Eshel and Heavens 2007, Paleoceanography, in press). The model dynamical core, tested extensively against observations in the northern Red Sea, performs well and tracks closely disparate observations while illuminating underlying physics. However, to be useful for studying land-ocean-atmosphere interactions in the western Sahel off the subtropical North Atlantic, the model must incorporate moist thermodynamics. To that end, a water vapor equation is needed, and it must include adequate representation of moisture sources and sinks. That is, precipitation parameterization is required. And while numerous parameterization schemes exist, they often require input variables that are absent from simplified, idealized models, like Eshel and Heavens'. Put differently, idealized models' highly truncated state vectors do not include as state variables the necessary input arguments into most precipitation schemes. As a result, such schemes cannot be used in the context of idealized models. Here we present a precipitation parameterization scheme for the Sahel region of subtropical western Africa. Sufficiently simple to be used in idealized models such as Eshel and Heavens', the scheme's predictions are surprisingly powerful, and compare extremely well with Sahel precipitation observations. In this talk, we present the scheme, test it against observations, and interpret its skill dynamically and physically.
H31E-0701
Intercomparison of the Performance of CLM3, NOAH, RUC, and STD Land Surface Schemes in the Weather and Research Forecasting Model
The Community Land Model version 3 (CLM3) developed by the National Center for Atmospheric Research (NCAR) was coupled into the Weather Research and Forecasting (WRF) Model version 2.2. The performance of WRF-CLM3 in predicting regional climate was quantitatively compared with that of WRF coupled to the soil thermal diffusion (STD), Rapid Update Cycle, and NOAH Land Surface Schemes. These land surface schemes represent a range of complexity within land-surface schemes. CLM3 is the most sophisticated model, with detailed snow and vegetation processes. The STD scheme is oversimplified, which only calculates soil temperature and neglects vegetation and snow physics. The RUC and NOAH schemes are intermediate in the detail, and the major deference between them is that RUC has a multi-layer snow scheme, and Noah has a single snow layer lumped with the topmost soil layer. WRF was driven by the National Centers for Environmental Prediction Reanalysis data II with each of these land surface schemes for one-year simulations over the period, 1 October 1995 to 30 September 1996, resulting in four one-year simulations for intercomparison. Each simulation has 30km-10km two-way nested domains. The 30 km domain includes the western U.S. and eastern Pacific, and the inner domain includes California and parts of Nevada, Oregon, and the eastern Pacific. Our analysis shows that WRF-CLM3 outperforms WRF-RUC, WRF-NOAH, and WRF-STD in simulating temperature and snow when compared with observations. The WRF-STD scheme, which does not include snow and vegetation processes resulted in the poorest results, with a dramatic overestimation of surface air temperature. However, regardless of the land surface scheme chosen, WRF reasonably well reproduces the winter precipitation, a major water resource for California, suggesting that the linkage between land surface processes and precipitation is not explicit. In general, land surface schemes play a significant role in the simulation of regional hydroclimate. Our coupled version of WRF-CLM3 has been shown to significantly improve WRF predictability.