A51E-01
Studies of Land Atmosphere Coupling Using the Land Information System
Recent studies have examined aspects of land atmosphere coupling including the roles of soil moisture and
vegetation, on the structure of the atmospheric boundary layer and initiation and evolution of clouds. However, due
to limits in computational resources and/or theoretical knowledge, many of these studies have utilized highly
parameterized representations of these components so that the true nature of land atmosphere coupling is still
unknown. The NASA/GSFC Land Information System (LIS; http:lis.gsfc.nasa.gov) has now been successfully
coupled to the Weather Research and Forecasting (WRF; http:www.wrfmodel.org) model, and now provides a
testbed for conducting studies of land-atmosphere coupling at water and energy cycle process resolving
horizontal spatial scales (1km or less). LIS is a high-performance Land Data Assimilation System (LDAS;
http:ldas.gsfc.nasa.gov) that encapsulates the capabilities of the North American LDAS (NLDAS) and the Global
LDAS (GLDAS) into a single software infrastructure. The original LIS consists of several land surface models
(e.g., Noah, CLM, VIC, HySSiB, Catchment) that can be run in two modes: uncoupled or coupled. In uncoupled
mode, the atmospheric boundary conditions are prescribed using observationally-based precipitation, radiation
and meteorological inputs, while in coupled mode, these inputs are predicted by the WRF model. In both cases,
LIS incorporates remotely sensed land surface parameters including Moderate Resolution Imaging
Spectroradiometer (MODIS)-based Leaf Area Index (LAI). In this talk, we will present results from various coupled
case studies indicating a strong sensitivity of the water and energy cycles to several controls, including soil
moisture, vegetation, and the atmospheric boundary layer. We will also demonstrate the value of remotely sensed
observations of ecosystem properties on predicting the timing and location of convection.
http:lis.gsfc.nasa.gov
A51E-02 INVITED
The role of land-atmospheric interactions in the initiation of deep convection for convection- resolving regional-scale models
Understanding and correctly modeling the feedback between land-surface variability and precipitation is important because of its potential benefit in improving weather and climate predictability. In summer, mesoscale boundaries play a critical role in the initiation of heavy precipitation. The zones of enhanced convergence along these boundaries have been recognized as areas of deep-convection initiation. The origin of these mesoscale boundaries includes synoptic-scale fronts, outflows from previous storms, orographic features, and differential surface heating. The differential heating can be enhanced by heterogeneities in land-surface conditions. The land surface may have differing impacts, depending on atmospheric conditions. Small-scale ground features, such as vegetation, hillslopes, and urban or industrial areas can also have subtle impacts that can determine the exact boundary and intensity of storms. We will review recent studies in employing advanced land surface models and high-resolution land data- assimilation system (HRLDAS) in convection-resolving models to investigate land-atmospheric interactions and their impacts on the initiation of deep convection. These studies include a flash flood case, a dryline convection case, and an 11-day heavy-precipitation episode. We found that fine-scale (L~10 km) boundary-layer circulations that directly trigger deep convection are confined within a mesoscale region containing a deeper and more unstable PBL, and that this region is a result of a surface sensible heat-flux maximum over dry soils. Results from these and other recent research studies provide some hope that the careful treatment of land-surface physics and soil moisture in convection-resolving models can lead to increased rainfall predictability. In particular, this should be achievable by improving 1) the representation of land surface processes, 2) the initialization of soil properties, and 3) the specification of various vegetation characteristics by combining modeling, new remote sensing capabilities, and data-assimilation techniques. It is, however, a more daunting challenge to incorporate the complex mesoscale interactions among the land surface, the boundary layer, and clouds, in large-scale climate models. These interactions play a critical role in determining the timing and location of convection initiation, and the intensity of precipitation, but operate on subgrid scales.
A51E-03
Does A Well-Calibrated Land-Surface Model Improve Numerical Weather Prediction ?
Accurate representation of heterogeneous land-surface energy and water fluxes is critical for the spatio-temporal evolution of planetary boundary layer and the predictability of summer-time deep cumulus convection in a numerical weather prediction (NWP) model. Although emergence of new satellite platform and in-situ data network has been improved the boundary conditions of land-surface models (LSMs), the performances of LSMs are uncertain on the regional scale. This is because LSMs have been generally tested and calibrated in limited time and spatial scales. Therefore, further improvements of LSMs require a regional-scale model calibration framework. We have established the regional-scale calibration framework over eastern U.S. for Unified Land Model (ULM) within the Goddard Land Information System (LIS). ULM is tuning-oriented structure that features common soil-vegetation-atmosphere-transfer parameterizations in different LSMs. This calibration framework iteratively reduces model-observation discrepancies in surface albedo, radiative temperature, and turbulent heat fluxes via Gauss-Marquardt Levenberg algorithm. The calibrated ULM significantly improves the representation of energy and mass fluxes over the eastern U.S. The well-calibrated ULM is tested for regional NWP by the Weather Research and Forecasting (WRF) coupled with Goddard LIS (WRF-LIS) platform. WRF-LIS model uses a grid spacing of 5km in a single nested domain forced by North American Regional Reanalysis on the lateral boundaries. This fine grid spacing together with the Goddard microphysics package can resolve convective precipitation without a subgrid convection parameterization. We will simulate several summer-time weather episodes over the eastern U.S. using the non-calibrated and well-calibrated ULM within the WRF-LIS platform, and will examine how the calibrated ULM improves the predictability of summer-time deep cumulus convection in comparison with the observations.
A51E-04
Storm Attributes and Cloud Resolution Model Performance in West Africa
Extensive datasets from the 2006 NASA African Monsoon Multidisciplinary Analyses (NAMMA) campaign offer fresh opportunities to investigate mesoscale convective systems (MCSs). For this presentation, we use rawinsonde profiles, surface fluxes, radar imagery (horizontal and vertical scans) , and precipitation data collected at the NAMMA field site in Kawsara, Senegal to investigate the temporal evolution of storms that occurred between August 19 and September 16, 2006. Two events, which took place on August 31 and September 7, are emphasized. The August 31st event crossed the coast around 0830 UTC and was of notable interest as this storm was associated with the development of Tropical Storm Gordon in the mid-Atlantic on September 10. The second storm of September 7th was also spatially extensive and organized as it passed over the West African coast around 0800 UTC. This storm did not lead to a tropical depression. A 2-dimensional version of the Goddard Cumulus Ensemble (GCE) cloud resolution model was initialized with rawinsonde profiles and energy fluxes from both of these case studies to learn how the model represents the growth and intensity of these storms. Model outputs were compared with the vertical reflectivity profiles and spatial distribution of rainfall. Results from this work will be discussed to identify new insights into MCSs and precipitation processes in West Africa, a region where surface-atmosphere interactions from continental and maritime environments are crucial for the development of storms. In addition, key results will be presented to illustrate how well the current configuration of the GCE model reproduces cloud formation and precipitating processes in West Africa.
A51E-05
Sensitivity of Clouds to Land Surface in CRM Simulations
A 20-day, continental midlatitude case is simulated with a three-dimensional (3D) cloud-resolving model (CRM) and compared to Atmospheric Radiation Measurement (ARM) data. Surface fluxes from ARM ground stations and a land data assimilation system are used to drive the CRM, respectively. The modeled cloud amount is compared with the observed, showing that the land data assimilation system has better surface fluxes than the ARM ones in summertime.
A51E-06
Dynamic LES Modeling of Observed Diurnal Cycles
The diurnally varying atmospheric boundary layers observed during the Wangara and CASES-99 field campaigns are simulated using the newly proposed locally-averaged scale-dependent dynamic subgrid-scale (SGS) model. This tuning-free SGS model enables one to dynamically compute the Smagorinsky coefficient and the subgrid- scale Prandtl number based on the local dynamics of the resolved velocity and temperature fields. We show that this SGS model-based large-eddy simulation (LES) has the ability to reproduce with great accuracy the characteristics of observed atmospheric boundary layers even with relatively coarse resolutions. In particular, the development, magnitude, and location of an observed nocturnal low-level jet are very well depicted. Some well- established empirical formulations (e.g., mixed layer and local scaling) are recovered with good accuracy by this LES-SGS parameterization. In this presentation, we also delineate the potential of this new-generation dynamic LES-SGS modeling approach to revise and improve existing numerical weather prediction boundary layer parameterizations.
A51E-07
Dynamics of Leaf Area for Atmospheric Models
We address the interactions between the terrestrial vegetation, surface hydrology, and the atmosphere. These interactions occur through many parameters and on multiple time scales. Biophysical coupling controls the evolution of the plant canopies and their impacts on weather and climate on time scales of days out to seasons. The concept of dynamic vegetation has been developed as a framework for modeling the longer time scale changes of the terrestrial biosphere that are of importance for climate models. This paper develops an approach to dynamic vegetation modeling on the time scales over which satellite observations observe variability of canopy properties and so allows linking a dynamic understanding of canopy processes with the observations that can be made to describe the dynamic state of the canopy and so to interface terrestrial remote sensing data with the dynamic state of weather and climate models. It builds on the idea that the most important short term drivers of natural canopy variability are water and thermal stresses.
A51E-08 INVITED
Sensitivity of the regional response to global warming associated with land cover
By now there is little doubt left about the fact that the increase in greenhouse gases (GHG) is producing global
warming. The question is whether the regional response to the GHG effect is uniform or depends on the land
characteristics and use. In this paper we show that the response is very dependent on the type of land cover and
use, and desertic and urban areas get more than their "fair share" of GHG warming, whereas broadleaf forested
areas have locally reduced warming.
We use the Observation minus Reanalysis (OMR) surface temperature trends method suggested by Kalnay and
Cai (Nature, 2003) to provide an estimate of the impact of surface effects on regional warming (or cooling). It
takes advantage of the insensitivity of the NCEP-NCAR Reanalysis (NNR) to land surface type, and eliminates the
natural variability due to changes in circulation (since they are also included in the reanalysis), thus separating
surface effects from greenhouse warming. Kalnay et al. (JGR, 2006) showed that over the US the OMR average is
small, but it has different regional signs, in good agreement with the regions of "urban heating and cooling"
obtained by Hansen et al (JGR 2001).
Lim et al. (GRL, 2005) compared two global observation-based data sets (CRU and GHCN) and two different
global reanalyses (NCEP-NCAR and ERA40) and MODIS-derived land classes. The results (Figure 3) showed
that the OMR trends have a strong dependence on the land-type, and that the OMR land-type dependence is
similar using either the NCEP-NCAR or the ERA-40 Reanalyses. Not unexpectedly, the ERA40 trends have about
half the amplitude, since this reanalysis uses air surface temperature observations indirectly (from an off-line OI
analysis of surface temperature) to initialize the soil temperature and moisture, so that the ERA40 surface
temperature are partially influenced by surface observations that are dependent on land surface properties. The
results show that OMR warming over barren areas is larger than most other land types, and that urban areas
show a large warming second only to barren areas. Croplands with agricultural activity show a larger warming
than natural broadleaf forests. The overall assessment indicates surface warming is larger for areas that are
barren, anthropogenically developed, or covered with needle-leaf forests. Lim et al (2006, submitted to JAMC)
extended this study to establish the dependence of OMR on NDVI (and hence on the Leaf Area Index, LAI), and
once again found very robust results. The results indicate that that OMR decreases with NDVI. Areas with little
vegetation suffer from warming higher than their "fair GHG share", whereas for highly vegetated zones, the OMR is
small or negative.
http:www.atmos.umd.edu/~ekalnay