H21I-01 INVITED
Land Surface Temperature Data Assimilation for Turbulent Flux Characterization
In situ surface turbulent flux measurements at sparse tower locations are not adequate to form mapping of these fields. Remote sensing instruments have the mapping capability but cannot sense flux directly. Nevertheless the diurnal march of land surface temperature (LST) and day-to-day anomalies in them do contain significant but indirect information on the relative magnitude of surface energy balance components. Model constraints are required to extract this information. We present a data assimilation scheme that combines LST data products from different sensors with different resolution, with a parsimonious land surface model in order to obtain estimates of surface fluxes, fluxes portioning and characteristics. The basic model has been upgraded in order to improve the evaporative fraction estimation over wet soil and vegetated areas. Mapping of evaporation fields are now possible even during days where satellite-derived LSTs are not available, e.g cloudy days. The assimilation scheme has been implemented and validated over SGP97 area and Italian regions. Results of the applications will be presented.
H21I-02 INVITED
Multi-scale assimilation of remotely sensed snow observations for hydrologic estimation
Data assimilation provides a framework for optimally merging model predictions and remote sensing observations of snow properties (snow cover extent, water equivalent, grain size, melt state), ideally overcoming limitations of both. A synthetic twin experiment is used to evaluate a data assimilation system that would ingest remotely sensed observations from passive microwave and visible wavelength sensors (brightness temperature and snow cover extent derived products, respectively) with the objective of estimating snow water equivalent. Two data assimilation techniques are used, the Ensemble Kalman filter and the Ensemble Multiscale Kalman filter (EnMKF). One of the challenges inherent in such a data assimilation system is the discrepancy in spatial scales between the different types of snow-related observations. The EnMKF represents the sample model error covariance with a tree that relates the system state variables at different locations and scales through a set of parent-child relationships. This provides an attractive framework to efficiently assimilate observations at different spatial scales. This study provides a first assessment of the feasibility of a system that would assimilate observations from multiple sensors (MODIS snow cover and AMSR-E brightness temperatures) and at different spatial scales for snow water equivalent estimation. The relative value of the different types of observations is examined. Additionally, the error characteristics of both model and observations are discussed.
H21I-03
Analysing soil moisture using the extended Kalman filter: Applications and results from ECMWF's global forecasting system
Soil moisture is one key parameter defining the turbulent fluxes between the land surface and the atmosphere. In Numerical Weather Prediction applications, soil moisture is often weakly constrained and used as a sink variable optimizing the turbulent fluxes and consequently the low level atmospheric temperature and humidity fields. However, future measurements from SMOS and SMAP will provide more direct observations of soil moisture, which will be used in combination with synoptic observations and soil moisture derived from ASCAT in the analyses. In this presentation we will introduce the new surface analysis system at ECMWF, which accommodates an extended Kalman Filter (EKF). The performance of the filter is compared against the current operational Optimal Interpolations System. It has been found that gain and soil moisture increments of the EKF are generally comparable with the ones obtained from the OI. However, there are local differences and changes in the vertical structure of the gain. The impact on atmospheric parameters and the forecast skill will be addressed. In the second part of the presentation results from a global demonstration study using ASCAT derived soil moisture in the EKF will be presented and compared against in-situ measurements.
H21I-04
Hillslope-scale soil moisture estimation with the ensemble Kalman Filter and a process ecohydrology model: Evaluation of anticipated microwave observations
Accurate knowledge of soil moisture at hillslope scales (e.g., 10's to 100's of meters) is critical to advancing hydrological applications such as irrigation scheduling, landslide prediction, wildfire fuel load assessment, and flood forecasting. Planned soil moisture remote sensing platforms, such as the European Space Agency's Soil Moisture and Ocean Salinity (SMOS) and the National Aeronautic and Space Agency's Soil Moisture Active-Passive (SMAP) missions, will provide global observation of soil moisture in the lower microwave region at frequent revisit intervals (2-3 days) and are partly targeted at improving soil moisture knowledge for applications. It is well recognized, however, that the data products provided by these missions are too coarse in spatial resolution to capture hillslope-scale variation in soil moisture. Process ecohydrology models are capable of simulating soil moisture at the spatial scales required, but suffer from uncertainties in the input data, model parameters and structure. Through a set of synthetic experiments, we assess the degree to which data assimilation through the ensemble Kalman Filter can be used to fuse simulated L-band microwave brightness and radar backscatter observations to uncertain hillslope-scale soil moisture estimates derived from a process ecohydrology model. We demonstrate that in a semiarid environment, assimilation of successive observations gradually improves the forecast soil moisture distribution, both in the near surface and the entire soil profile. Representing the role of topography in controlling moisture redistribution, a measurement equation system that accounts for topographic impacts on observing geometry, and adequate characterization of uncertainty in soil hydraulic and thermal properties are critical to the success of a hillslope-scale soil moisture data assimilation system. While this work suggests data assimilation is potentially useful for improving knowledge of soil moisture at hillslope scales, additional studies are required to evaluate more complex hydrologic scenarios and settings, as well as other data assimilation algorithms.
H21I-05 INVITED
Deriving Monin-Obukhov Similarity Functions from Dynamic Large-Eddy Simulations
The temperature structure parameter (CT2) and the turbulence kinetic energy dissipation rate (ε) are the primary turbulent variables utilized in scintillometry. Monin-Obukhov (M-O) similarity functions are used to estimate heat and momentum fluxes from CT2 and ε. These M-O similarity functions are empirically derived from fast-response turbulence observations collected during different field campaigns. The atmospheric boundary layer field measurements are seldom free from mesoscale disturbances, wave activities, and nonstationarities. Thus, it is not surprising that quite a few contending similarity functions have been proposed in the recent literature. While using these traditional M-O similarity functions in scintillometry, one faces two fundamental problems. First of all, the application of point-measurements-derived similarity functions in scintillometry (where fluxes are estimated over an area) is conceptually flawed. Secondly, the estimation of CT2 and ε from time-series data require the assumption of Taylor's hypothesis. One could circumvent both these difficulties by deriving the similarity functions from a database of high-resolution dynamic (tuning-free) large-eddy simulation fields. The pros and cons of this alternative approach will be the focus of this presentation.
H21I-06 INVITED
New Mexico Scintillometer Network in Support of Remote Sensing and Hydrologic Modeling
In New Mexico, a first-of-its-kind network of seven Large Aperture Scintillometer (LAS) sites was established in 2006 to measure sensible heat fluxes over irrigated fields, riparian areas, deserts, lava flows, and mountain highlands. Wireless networking infrastructure and auxiliary meteorological measurements facilitate real-time data assimilation. LAS measurements are advantageous in that they vastly exceed the footprint size of commonly used ground measurements of sensible and latent heat fluxes (~100 m2), matching the pixel- size of satellite images or grid cells of hydrologic and meteorological models (~0.1-5 km2). Consequently, the LAS measurements can be used to validate, calibrate, and force hydrologic, remote sensing, and weather forecast models. Results are presented for: (1) variability and error of sensible heat flux measurements by scintillometers over heterogeneous terrain and (2) the validation of the Surface Energy Balance Algorithm for Land (SEBAL) applied to MODIS satellite imagery. Findings from this study are discussed in the context of researchers' and practitioners' data assimilation needs.
H21I-07
On the use of Land and Ground Data to Improve Integrated Turbulent Fluxes Estimated From Scintillometer Measurements
Scintillometry has been validated in many contexts of homogeneous and patchy land covers. Nevertheless
natural landscapes are often more heterogeneous. This disables classical validation methodology which
consists in comparison with aggregated Eddy correlation data.
This study propose a simultaneous analysis of energy and water budget at hourly and daily scale to improve
estimation of turbulent fluxes at scintillometer scale over a very heterogeneous area.
Within the framework of the "AMMA-CATCH" program, a Large Aperture Scintillometer (LAS) has been
installed in a small catchment (12 km2), located in the north Benin, a region exposed to soudano-sahelian
climate. The site is instrumented with several ground stations to monitor the vadose zone, water table,
precipitations and river discharge. Two case studies will be presented. The first one covers the end of the dry
season. During this period two isolated rainfalls occurred which give a unique opportunity to study energy
and water budgets simultaneously. The second one deals with the rainy season and offers to characterize
evapotranspiration when vegetation suffers no water shortage.
First, a methodology is proposed to aggregate surface parameters z0 and d, which are used to
compute sensible heat flux from LAS data. Sensible heat fluxes from LAS and EC data are then compared.
This shows a relatively good agreement where the scattering is mainly attributed to footprint variability. A
relevant hourly residual latent heat flux is then obtained through the energy balance equation with careful
attention brought to the aggregated net radiation and ground heat fluxes.
The residual of the energy budget equation is then compared to soil water losses from vadose zone and
water table in order to evaluate whether this estimation is consistent with the water budget of the ground. For
the dry season period, the dynamic of ground water losses and evapotranspiration are well correlated.
During the wet season we find a constant evaporative fraction which is relevant when vegetation runs into
good transpiration conditions. Finally, this study shows how combined energy and water budget analysis can
help to better understand water transfers at watershed scale.
http://ltheln21.hmg.inpg.fr/catch/?page=homepage&lang=en
H21I-08
Multi-scale Modeling of Energy Balance Fluxes in a Dense Tamarisk Riparian Forest
Remote sensing of energy balance fluxes has become operationally more viable over the last 10 years with the development of more robust multi-layer models and the availability of quasi-real time satellite imagery from most sensors. Riparian corridors in semi-arid and arid areas present a challenge to satellite based techniques for estimating evapotranspiration due to issues of scale and pixel resolution, especially when using the thermal infrared bands. This paper will present energy balance measurement and modeling results over a Salt Cedar (Tamarix Ramosissima) forest in the Cibola National Wildlife Refuge along the Colorado River south of Blythe, CA. The research site encompasses a 600 hectare area populated by mostly Tamarisk stands of varying density. Three Bowen ratio systems are installed on tall towers within varying densities of forest cover in the upwind footprint and growing under varying depths to the water table. An additional eddy covariance tower is installed alongside a Bowen ratio system on one of the towers. Flux data has been gathered continuously since early 2007. In the summer of 2007, a Scintec large aperture scintillometer was installed between two of the towers over 1 km apart and has been working continuously along with the flux towers. Two intensive field campaigns were organized in June 2007 and May 2008 to coincide with LANDSAT TM5, MODIS and ASTER overpasses. High resolution multispectral and thermal imagery was acquired at the same time with the USU airborne system to provide information for the up- scaling of the energy balance fluxes from tower to satellite scales. The paper will present comparisons between the different energy balance measuring techniques under the highly advective conditions of the experimental site, concentrating on the scintillometer data. Preliminary results of remotely sensed modeling of the fluxes at different scales and model complexity will also be presented.