H11H-01 08:00h
Determination of the Infrared Emissivity with Multi-spectral Thermal Infrared Data from Space
Knowledge of the land surface emissivity is important for estimating the longwave radiation budget, a decrease of soil emissivity by 0.1 will increase ground and air temperature by about 1.1 C and 0.8C and decrease net and upward longwave radiation by about 6.6 and 8.1 W/m*m, respectively. The multi-spectral thermal infrared data from the Advanced Spaceborne Thermal Emission and Reflection (ASTER) radiometer provides a new tool for observing land surface emissivity . ASTER has 5 channels in the 8 to 12 micrometer wave band with 90 meter resolution. These data can be used to assess the spectral and spatial variations of surface emissivity when used with the Temperature Emissivity Separation (TES) algorithm. TES makes use of an empirical relation between the range of observed emissivities and their minimum value to extract the temperature and 5 emissivities from the 5 channels of ASTER data. The approach was validated with ASTER data acquired over the Jornada Experimental Range and the White Sands National Monument in New Mexico between 2001 and 2003 yielding good agreement with ground measures of emissivity. The approach was extended to produce maps of emissivities over a 400 x 1200 km area for a desert region of North Africa, including the sand dunes of the Grand Erg Oriental using data acquired in 2001 and 2002. The spectra for the sand dunes showed good agreement with that expected for quartz sand based on laboratory and field measurements. A multiple regression approach was used to relate the emissivities of the 5 ASTER channels to the window channel emissivity. The results were compared with a classification based emissivity map and significant differences were found, ranging between -0.08 and +0.06. The spatial variation of the emissivity observed by ASTER is from 0.8 to 1, which corresponds to a range of 15 w/m*m in the net surface longwave radiation under a dry atmosphere. These results show that ASTER data can be used to map the spatial and spectral variations of surface emissivity over large areas in particular the deserts of the world for which there is much exposed soil and sand. To extend the map to continental scales a relationship between (a) the ASTER broadband emissivity map and (b) spectral emissivity and spectral reflectance data from MODIS data was developed. We applied this regression to MODIS data and generated a broadband emissivity map for North Africa. The range of the broadband emissivity was found to be between 0.86 and 0.96 for the desert area. The expected RMS error of the map is about 0.02. Such an emissivity map has been used as an input to a climate model and improves the prediction of surface and air temperatures by up to 1 degree C.
H11H-02 08:15h
Streamflow Predictions using GOES Solar Radiation Based Evapotranspiration in the Sacramento Soil Moisture Accounting Model
The National Weather Service River Forecast System (NWSRFS) uses the Sacramento Soil Moisture Accounting (SAC-SMA) model to generate river and flood forecasts across the U.S. One of the required inputs needed to run SAC-SMA is potential evapotranspiration (PET), which the NWSRFS computes using the Penman combination equation in the Synoptic Data Transfer (SYNTRAN) utility program. Of the hydrometeorologic forcing data required to compute PET, incoming solar radiation data availability is the greatest challenge for operational hydrologic applications. Total sky cover observations, which SYNTRAN uses to estimate incoming solar radiation, have been phased out of existence with the implementation of the Automated Surface Observation System (ASOS). Therefore, NWS needs to find solar radiation data sets to replace the SYNTRAN solar radiation data without diminishing SAC-SMA model accuracy during streamflow simulations. We use Geostationary Operational Environmental Satellite (GOES) gridded solar radiation and meteorological forcing data from the Distributed Model Intercomparison Project (DMIP) test basins to calculate the PET estimates needed to simulate streamflow with the SAC-SMA model. Resultant daily streamflow predictions are compared with observed streamflow data.
H11H-03 08:30h
Quantifying the Spatial Distribution of Evapotranspiration over Canada With a Process Model Using Remote Sensing, Meteorological, and Soil Data
The evapotranspiration (ET) from all Canadian landmass is estimated at daily steps and 1 km resolution using a process model named Boreal Ecosystem Productivity Simulator (BEPS). The model is driven by remotely sensed leaf area index and land cover maps, as well as soil water holding capacity and daily meteorological data. All the major ET components are considered: transpiration from vegetation, evaporation of canopy-intercepted rainfall, evaporation from soil, sublimation of snow in winter and in permafrost and glacier areas, and sublimation of canopy-intercepted snow. In forested areas, the transpiration from both the overstory and understory vegetation is modelled separately. The Penman-Monteith method was applied to sunlit and shaded leaf groups individually in modelling the canopy-level transpiration, a methodological improvement necessary for forest canopies with considerable foliage clumping. The modelled ET map displays pronounced east-west and north-south gradients as well as detailed variations with cover types and vegetation density. It is estimated that, for a relative wet year of 1996, the total ET from all Canada's landmass (excluding inland waters) was 2037 km3. If compared with the total precipitation of 5351 km3 based on the data from a medium range meteorological forecast model, the ratio of ET to precipitation was 38 %. The ET averaged over Canadian land surface was 228 mm/yr in 1996, partitioned into transpiration of 102 mm/yr and evaporation and sublimation of 126 mm/yr. Forested areas contributed the largest fraction of the total national ET at 59 %. Averaged for all cover types, transpiration accounted for 45 % of the total ET, while in forested areas, transpiration was contributed 51 % of ET. Modelled results of daily ET are compared with eddy covariance measurements at three forested sites with a r2 value of 0.61 and a root mean square error of 0.7 mm/day.
H11H-04 08:45h
Evaluating the Influence of Various Vegetation and Soil Types on Water and Energy Balances in Northern Wisconsin Using a Dynamic Biosphere Model
Variations in vegetation composition and soil types across large landscapes significantly influence coupled energy-water-carbon cycles at the watershed scale. Past efforts to quantify northern Wisconsin's energy and water balances using atmospheric, groundwater and other landscape models have used only one land cover type. The next step is to account for the heterogeneity that is found in both soil and vegetation across the region. Therefore, mechanistic models that account for land cover changes more completely are important to investigating this landscape. This study uses a process-based terrestrial model, Integrated Biosphere Simulator (IBIS), to compare evapotranspiration, sensible heat, soil moisture, surface runoff, and drainage over various land covers including grasslands, shrublands, coniferous and deciduous forests, as well as soil types ranging from sand to clay. The temporal sensitivities of land cover are evaluated from 1951-2000, comparing long-term averages, innerannual variability and seasonal cycles with attention given to wet and dry years and extreme precipitation events. Results show 1) when compared to observations at the 447-m WLEF-TV flux tower, the model successfully simulates variation in monthly latent and sensible heat flux and 2) hydrological sensitivities of water and energy budgets components vary depending on timescales; evapotranspiration, for instance, is less variable than drainage seasonally, but has a greater sensitivity to land cover change over the long-term.
H11H-05 09:00h
Simulation of vegetation, soil characteristics, and topography effects on soil water distribution and streamflow timing over a semi-arid mountain catchment
Soil water ($\theta$, m$^{3}$m$^{-3}$) and soil characteristics act as intermediaries, along with plants and climate, modifying and modulating streamflow timing and quantity-the majority in the intermountain US west resulting from spring-melt events of accumulated winter snow. The antecedent soil water conditions also predispose different patterns and dynamic responses, especially in semiarid, mountain regions. The context of soil water, analyzed using modeling, is necessary to describe the processes of soil water dynamics. In this research, two years of neutron probe soil water data from a small, semiarid mountain catchment were evaluated using a vertical flow, combined snowmelt-soil water, capacitance-parameter model with available snowmelt data and climate data as driving inputs. Model parameters were vegetation characteristics and soil properties. Results at the point scale show good fit at many locations while a few have poor simulation results at depth. The discrepancies are hypothesized to be due to lack of understanding of parameters such as rooting depth of trees; heterogeneity of parameters within the soil layers; using capacitance parameters that treat some variables as constants; exclusion of lateral flow processes that must occur in some locations due to basin geometry and nature of soil-fractured bedrock interface; and rising water table effects that can be seen in the gleying of clayey soils near drainage lines. Driving parameters were then distributed over the 0.36 km$^{2}$ catchment using the regional 10 m DEM, soil maps, remotely sensed color-infrared imagery, and the spatiotemporal distributions of soil water from previous research. The model was run discretely at each pixel. Results matched point data simulations well. Simulated throughflow, totaled over the watershed, compared well with weir measured streamflow in timing and quantity indicating accurate representation of parameters over the watershed, proper calibration, and well described processes.
H11H-06 09:15h
Changes in Vegetation and Surface Fluxes Associated with the North American Monsoon in 2004
Observations show that the onset of summer rains occurs in late June over northwestern Mexico, and that the greatest rainfall over the foothills of the Sierra Madre Occidental occurs in July. The region of maximum rainfall migrates oceanward with time such that by September the greatest rainfall is along the coastal region, with noticeably less rainfall inland. The onset of the rains is accompanied by a rapid leafout of the vegetation with an associated increase in evapotranspiration. We hypothesize that the rapid increase in foliage, quantified by the leaf area index, has a feedback on the atmospheric radiation balance in the surface layer which in turn affects the maximum temperature of the inland terrain. Thus the boundary layer undergoes a rapid transition from a deep well-mixed dry layer during the pre-monsoon period to a shallower, but much more humid layer during the post-onset. The changes in the vegetation characteristics during the onset period, together with the seasonal increase in the Gulf of California SSTs contribute to a seasonal march of the region of maximum precipitation towards the coastal areas and away from the mountain slopes as the summer progresses. The North American Monsoon Experiment (NAME) provides an excellent opportunity to study this problem. During the Enhanced Observation Period in the summer of 2004, the existing operational networks of upper air soundings and precipitation were significantly improved for the core monsoon region. In addition 3 micrometeorological towers were installed to provide information about the surface fluxes throughout the monsoon development. Preliminary results will be presented and related to changes in vegetation which were monitored using MODIS satellite images.
H11H-07 09:30h
Landscape Controls on Monsoon Soil Moisture Distribution in Northern Sonora, Mexico
It is widely known that topographic and ecosystem variability exert a strong signature on the surface soil moisture distribution over regional scales, yet there is little observational evidence available to quantify this relationship, in particular over arid and semiarid areas. To properly determine the topographic and vegetation effects on soil moisture, and potentially use terrain and land-cover data to downscale remote sensing data, a spatially-extensive field sampling of soil moisture in regions of topographic variability is required. The SMEX04-NAME field campaign in northern Sonora, Mexico provides a unique opportunity for understanding the linkages between topographic position and the statistical properties of soil moisture during the North American Monsoon season in a sparsely observed region. In this study, we present a preliminary analysis of the data collected from both an in-situ network of soil moisture sensors and a two-week field campaign over a west-to-east transect in mountainous northern Sonora. The time evolution of topographic profiles of observed hydrometeorological quantities (rainfall, relative humidity, air temperature, pressure) and surface hydrologic properties (soil moisture, soil temperature) will be presented. We utilize available spatial data sets on the regional topography, soil types and vegetation distributions to quantify the linkage between monsoonal soil moisture, topography and ecosystem pattern. A preliminary comparison will also be made between hydrometeorological and surface hydrologic data and terrain attributes (elevation, slope, aspect, curvature) within a watershed encompassing the study transect. The topographic and vegetation data sets will then be used to interpolate field data and create soil moisture maps that can be aggregated to the aircraft and satellite footprint resolutions for conducting areal average comparisons to remotely sensed observations.
H11H-08 09:45h
Should We Be Concerned about Second Order Approximations in Hydrologic Models?
Hydrologic models used in land-atmosphere interaction studies have seen significant grown in sophistication over the years. However, soil-moisture transport equations remain rather simplistic. There is a growing trend of implementing these models at ever increasing resolutions. This necessitates the need to incorporate processes that arise at smaller scales but could be neglected for larger scales studies. This presentation will show that the soil-moisture transport can be represented as a scale dependent function that uses second order approximations for parameterization. These parameterizations incorporate properties such as the dependence of soil-moisture on the sub-grid statistics of topographic attributes (slopes and curvatures). These statistics change with the resolution of the model, and therefore, these formulations allow better characterization of vertical and lateral soil-moisture transport as model resolution changes.