Research Elements

Global Scale Prediction and Documentation

GCM Simulation of Hydrologic Cycle Components

Barron and Jenkins completed a series of control experiments using GENESIS for the EOS project, including an AMIP simulation with specified seasonally varying sea surface temperatures (SSTs) from 1979 to 1988, SST climatologies, a mixed layer ocean experiment, and R15 and T42 versions. Barron and Jenkins (1994) assessed the GENESIS precipitation over the United States for different GCM resolution and ocean specifications, noting that the ocean component (mixed layer versus longitudinal variation in SST) was more significant in predicting U.S. precipitation than was the difference between R15 and T42. GCM predictions were found to be inad equate for regional climate evaluations. A detailed comparison of the results of these experiments is in preparation.

Future research plans include intercomparison of CCM2, advanced GENESIS and UK Meteorological Office GCM predictions of precipitation, humidity and temperature structure, comparison of GCM observations with EOS Pathfinder data and GEWEX global precipitation climatology and completion of base runs for nested model experiments, focussed neural net analysis of the Susquehanna River Basin, and incorporation of improved cloud parameterizations in GENESIS.

GCM Sensitivity to Lower Boundary Forcing

Many of the experiments described under Objective 1 for different ocean specifications (AMIP SSTs, climatological SSTs, mixed layer model) and the results (Barron and Jenkins, 1994) have direct bearing on this goal as well. Recent experiments at T31 resolution (Robertson et al., 1995a) and the studies cited above indicate: 1) that tropical SST forcing has a robust quasi-linear response in the vicinity of the heating - - rainfall, divergent circulations and radiative anomalies for two El Ninos are well captured by the model, 2) the middle latitude response is characterized by a much lower signal to noise ratio, 3) warm season, mid-latitude anomalies such as the 1988 drought over the U.S. show significant predictability, and 4) in agreement with ERBE results, comparison of 1987 and 1988 greenhouse forcing calculations show some negative water vapor feedback over the tropical Pacific associated with elevated SSTs and deep convection.

In addition, Dutton and Barron (1995) have completed a series of experiments for changes in GENESIS land surface vegetation with two purposes, to determine the significance of vegetation factors in climate and to assess the potential significance of vegetation changes for future cli mate. These experiments demonstrate that the greatest sensitivity occurs for changes between low level vegetation (grasslands and tundra) and multi-storied vegetation, and between ever green and deciduous forests. Large regional surface temperature changes (greater than 5 degrees C) and global changes of almost 1 degree C were noted in association with changes in the distribution of grasses, tundra and forests. Each of these vegetation types is sensitive to climate change.

Future research will include continued: (1) Use of EOS Pathfinders, including SSM/I retrievals of vapor, liquid water, precipitating ice, and surface wind stress and MSU radiances for examining the hydrologic cycle and its sensitivity, (2) Use of the SSMT/2 water vapor profiling capability to quantify water vapor feedback, (3) Use of ISLSCP data, with observed precipitation, and satellite estimates of surface incident solar radiation to synthesize global soil moisture time series, (4) Completion of the vegetation experiments for different vegetation classes, leading to a more detailed analysis of the significance of the vegetation parameters, including the e-folding time for snow cover, leaf abscission, and the rooting depth of the vegetation. These experiments will allow a better assessment of the uncertainties and the significance of specific variables in governing climate-vegetation sensitivity.

Global Data Sets for Documentation of Global Change

Results reported in Robertson and McCaul (1994), Barrett et al. (1994) and Cohen and Robertson (1994) have focussed on development of a gridded (2.5 degree resolution, 2 x daily) SSM/I precipitable data product and a unique semi-prognostic approach for assimilating remotely sensed moisture data. By combining SSM/I water vapor and ECMWF large scale vertical motions we have derived conservation equations which link vapor, liquid and ice. As one important demonstration of this methodology, a consistent vapor and condensate budget for the TOGA-COARE period has been developed which connects the large-scale divergent circulation to atmospheric hydrologic processes. These results have a variety of uses, from providing a 3-D time structure of atmospheric moisture constituents, to serving as testbeds for convective and cloud parameterizations, and calibrating microphysical parameterizations which influence the upper tropospheric water budget. The products are an important contribution to EOS 4-D data as similation efforts. Additional analyses focussed on examination of the interannual variability and persistence of soil moisture anomalies in the global model (Fitzjarrald et al., 1994; 1995).

The analysis of MSU data continues to provide high precision data for climate studies. To date, over 25 refereed papers have been published for which the MSU data were the sole or main source of information for the scientific results. Improvements have been applied as intersatellite variations have been analyzed due to orbital drift, instrument dynamic range, absolute biases and missing data. These data are heavily used in assessments of global climate change. Key results are (1) proof that sequentially-launched polar orbiting satellites with microwave instruments designed for weather observations provide sufficient precision to create long-term data sets, (2) the global temperature of the troposphere has declined at a rate of 0.06 degrees per decade since 1979, however removal of the effects of El Nino and volcanoes results in an upward trend of 0.09 degrees per decade since 1979. Latest results of transient climate model runs (including aerosols) show similar trends, (3) the last two years have exhibited the coldest lower stratospheric temperatures on record. A key part of our global climate model simulations is the fidelity with which the GENESIS GCM reproduces the actual global temperature distribution and variability. MSU data for 17 years (temperature and oceanic precipitation) are now available for the variability studies. Example publications and conference reports which are contributions from this investigation for 1994 and 1995 include (Christy, 1995; Christy et al., 1995a,b,c; Balling and Christy, 1995; Christy and Drouilhet, 1994; Christy and McNider, 1994; Christy and Goodridge 1995; Spencer et al., 1994; Karl et al., 1994).

Future research emphasis will be on validation of GCM simulations. The inadvertent positional drift of NOAA-11 has opened up the possibility of identifying the diurnal cycle of tropospheric and stratospheric temperatures. Effort will also be focussed on predictive studies on the 10-day to seasonal time scale.

Cloud and Precipitation Processes

We have characterized the structure of the marine boundary layer (FIRE, ASTEX, TIWE) and fractional cloudiness (laser ceilometers) and continental stratus (using cloud radar, lidar, ceilometers, microwave radiometers) in order to evaluate and develop cloud parameterizations. Major accomplishments include: (1) development of a hierarchy of techniques to drive cloud parameterizations with data, separately from GCM studies, in order to verify GCM cloud and radiation predictions (Mace and Ackerman, 1995, Mace et al. 1994). This method allows integration across the scales of GCMs and mesoscale models, which are central to this EOS investigation. (2) demonstrated the feasibility of using a simple boundary layer model to provide a regional simulation of boundary layer structure and fractional cloudiness (with S. Wang and P. Minnis) (Wang,1994a,b, Wang and Wang, 1994). (3) Modified Tiedke's prognostic cloud scheme and implemented it in a two-layer boundary-layer model (Wang, 1995). Coupled with a mass flux convective mixing scheme this technique is a good framework for parameterization of marine boundary layer clouds in GCMs. (4) Defined the structure of the marine boundary layer over the past several years over the tropics and subtropics during major field programs (Albrecht et al., 1995a,b), developing composite soundings which were then used to evaluate model parameterizations that represent fractional cloudiness as a function of boundary layer structure. Parameterizations based on the relative humidity in the cloud layer agree better with observations than those that account for cloud-top entrainment instability on fractional cloudiness. These composite data sets are being used to test the boundary layer regional model for this EOS project, and support the atmospheric and cloud-related products generated from MODIS and CERES. Additional contributions included elucidation of cloud structure using radar (Syrett et al., 1994).

Increased emphasis in future research will be placed on continental boundary layer clouds as characterized using surface-based remote sensors, aircraft and satellites with the goal of realistically simulating continental temperatures and annual and diurnal temperature cycles. This will be based on observations collected during the fall of 1994 and the spring of 1995. We plan to compare the cloud structure observed from satellites (in cooperation with P. Minnis) with cloud characteristics obtained from surface based remote sensing, providing an important framework for utilizing EOS cloud products to infer the characteristics of continental boundary layer clouds.


Table of Contents Previous Section Next Section Download Report
Graphics and HTML text markup by Phil Kolb
Comments to: R. White, raw@essc.psu.edu