SATELLITE OBSERVATIONS

We will use CERES, MODIS, ASTER and AIRS/AMSU/HSB satellite data from Terra and Aqua platforms and GMS5 data for a time period of one year (June 2004 – June 2005) in the proposed research.  The CERES data will be used to examine differences in surface and top-of-the-atmosphere radiation budgets.  The land surface data from MODIS, crucial inputs for the soil vegetation model within RAMS, will be used to construct numerical modeling simulations for current and pristine land use scenarios.  Differences in cloud formation and microphysics between agricultural and native vegetation areas will be explored using MODIS and GMS5 data.  MODIS and ASTER data will be used to examine surface energy, moisture fluxes and soil moisture as a function of land use.  We will use TRMM data to examine precipitation associated with cloud systems influenced by land use.  MODIS retrieved cloud particle size will be used to detect the presence of drizzle in cumulus clouds.  Satellite retrieved soil moisture will also be used to initialize the soil model in the RAMS.  We will use AIRS/AMSU/HSB satellite data to study differences in atmospheric thermodynamic profiles over regions of contrasting land use.

We will use CERES data to examine the effect of land use on top-of-the -atmosphere and surface clear sky radiation budgets.  CERES products include both reflected solar and Earth emitted radiation at the top-of-the-atmosphere.  Surface shortwave and longwave fluxes will be derived from the CERES top-of-the-atmosphere observations using Langley parameterized shortwave/longwave algorithms (Gupta et al, 1992; Gupta et al., 2001).  These algorithms use atmospheric temperature and humidity profiles derived from European Center for Medium Range Weather Forecasts (ECMWF) data and MODIS cloud masks created using the Interactive Visualizer and Image Classifier for Satellites (IVICS) software.  A monthly global emissivity database is used to provide the emissivity data for the model.  The seasonal variation of radiative energy flux components in combination with seasonal variation of NDVI derived from MODIS data will be used to quantify the effect of agriculture practices on the top-of-the-atmosphere and surface radiation budgets of this area.

  We will use MODIS and ASTER satellite data for retrieval of surface energy and moisture fluxes.  The MODIS data, available twice daily, will be used to examine seasonal the variation of surface energy budget and soil moisture availability at 1km spatial resolution.  ASTER data is available over the same location approximately every 16 days, and surface energy, moisture fluxes and soil moisture availability can be retrieved at 90m spatial resolution.  We will use the triangle method of Gillies et al. (1997) for satellite retrievals of surface energy and moisture fluxes.  Our preliminary work suggests that this technique, which has been validated over other geographical areas, performs adequately in the proposed study region (Ray et al., 2003).  However, we plan to conduct a rigorous validation of this technique by comparison against the in situ surface flux and soil moisture measurements.  We will compare point in situ measurements of sensible heat, latent heat flux and soil moisture availability from the Bowen ratio system and the flux tower  against corresponding values retrieved from co-located ASTER data.  The averaged values of energy, moisture fluxes and soil moisture availability retrieved from ASTER pixels located within a 1km MODIS footprint will be compared against the MODIS retrievals for that pixel.
We will also examine the feasibility of developing parameterizations for retrieving surface sensible heat and latent heat fluxes from CERES data. The proposed parameterization will use CERES derived surface radiative fluxes in combination with NDVI values derived from MODIS data, the soil vegetation atmosphere transfer model and atmospheric profile information from outputs of forecast models such as ECMWF.  We will utilize in situ observations from both the Bowen ratio station and the Sturt Meadows flux tower for developing this parameterization. A realistic parameterization could be a very powerful tool for creating global datasets of satellite derived energy fluxes and examining the land use climate interactions.

In addition to in situ observations obtained from radiosonde launches by the Australian meteorological bureau and proposed field campaigns, we will also use temperature and humidity profiles derived from AIRS/AMSU/HSB data.  This data will be used to explore the differences in atmospheric conditions as a function of land use.  At 50km spatial resolution, AIRS/AMSU/HSB data will provide significantly more detailed information on the spatial variability of atmospheric conditions compared to the coarse spatial resolution radiosonde network.  We will use this data to compare the seasonal variation of boundary layer profiles as a function of land use, which will be useful for understanding the differences in cloudiness between agricultural areas and native vegetation.  This data will also provide quantitative measures of horizontal thermal gradient in the vicinity of the bunny fence, which is an important factor when considering the land surface heterogeneity effects.

One of the main goals of this study is to explore how land use and landscape heterogeneity influence cloud formation and impact regional hydrology.  For this purpose, we will use meteorological and satellite data to identify situations where land use and landscape heterogeneity may be influencing cloud formation.  Examples of such situations will include those days in summer when clouds form mostly over native vegetation, days in winter when this behavior is reversed with clouds forming over the agricultural area  and when mesoscale convective systems form almost directly over the bunny fence region.  The first two categories result from variations in boundary layer development between agricultural and native vegetation areas.  The landscape heterogeneity is the potential cause for the third category situation.  We will quantify how often such situations occur and also the amount of precipitation associated with the cloud systems.  This will enable us to observationally quantify the effect of land use on the local hydrology.  We will also complement these estimates with model simulated precipitation for current and pristine land use scenarios.

For a few selected cases belonging to each of the above listed categories, we will use numerical modeling simulations with current and pristine land use scenarios to verify if the nature of land use is responsible for the observed behavior.  Along with numerical modeling simulations, we will examine the atmospheric thermodynamic profiles, surface meteorological data, and surface energy budgets associated with these situations in order to understand the interaction processes between land use and climate.
We will use MODIS and Japanese GMS5 data to examine the cloud formation in this area.  Daytime GMS5 data for this area is available from an archive at Kochi University in Japan. Prof. Tokio Kikuchi, who is a collaborator in the proposed study, has agreed to provide hourly daytime GMS5 data for the study region at no extra cost to the project.  Once we identify days belonging to above described categories, we will use in situ rain gauge and satellite observations to quantify precipitation amounts.  We will use TRMM satellite observations to quantify the precipitation amounts.  For cumulus cloud fields, we will also use MODIS derived effective cloud particle size to detect the presence of drizzle.  The presence of drizzle will be used to determine the percentage of clouds with the potential of producing precipitation.

We will also use MODIS-derived cloud effective radii and cloud liquid water path values to examine the differences in cloud microphysics between native vegetation and agricultural areas.  Our prior observational and numerical modeling studies suggest that cloud characteristics are influenced by land use (Lawton et al., 2002; Nair et al., 2003).  We will use MODIS derived cloud microphysics to explore the differences in cloud characteristics between native vegetation and agricultural areas.  Atmospheric thermodynamic profiles, surface energy, moisture fluxes and soil moisture will be used to analyze the processes through which land use influence s cloud characteristics.

In the Southwestern coastal regions of Australia, along with areas cleared for cultivation, there are also patches of land that still contain native vegetation.  We will use these adjacent areas of contrasting land use along the coast to explore the impact of land use on sea breeze circulations.  We will use clouds that form along the sea breeze fronts as a tracer for these circulations and examine how they differ over native vegetation and croplands using MODIS and GMS5 satellite data.  We will also quantify the amount of precipitation associated with these systems using in situ observations and also using TRMM satellite data.

GMS5 satellite imagery will be supplied at no cost to the proposed project by Prof. Tokio Kikuchi, from Kochi University in Japan hourly from 0800 to 1500 LST for the duration of the project.  The Structural Thresholding method of cumulus cloud detection will be applied to determine the location of cumulus clouds (Nair et al, 1999; Ray et al, 2003).  Then scenes will be analyzed to determine patterns of cloud formation over the agricultural and natural vegetation regions, not only on the two sides of the bunny fence, but also in coastal regions.  Finally, monthly frequency of occurrence of cumulus clouds will be computed from 0800 to 1500 LST at hourly intervals.  We have already conducted this type of analysis for the time period of 1999 to 2000, reported in Ray et al (2003), who show both hourly maps for September and December, 2000, as well as monthly maps at 1300 LST for 2000.

<--top