C4 - A Mid Course Summary

Why Study Clouds? | Ongoing Research Plan | Education | Knowledge Transfer | Summary

Why Study Clouds?

The effect of clouds on climate is one of the principal sources of uncertainty in our ability to predict future climate change and is thus a top priority in climate research. The Center for Clouds, Chemistry and Climate (C4) is dedicated to reducing the uncertain ties involved in cloud-chemistry-climate interactions, thereby enhancing our understand ing of the climate system, and improving our ability to predict how it will respond to human and other influences.

CEPEX: A Major C4 Achievement

C4 seeks to develop a theoretical, observational, and modeling base for understanding and predicting cloud-chemistry-climate interactions. To this end, C4 took the lead in designing, implementing, and analyzing the results of a major field observing program known as CEPEX the Central Equatorial Pacific Experiment, conducted in March of 1993. The project yielded important new information about the absorption of solar radiation by clouds and the relative roles of cirrus cloud radiative effects and surface evaporation in limiting maximum sea surface temperatures in the equatorial Pacific. This large, interdisciplinary experiment, involving 15 institutions from the US and Germany, has created a much-needed observational basis for unraveling phenomena which were previously understood only theoretically. C4's infrastructure and frame work for cooperative research were instrumental in making CEPEX a success.

In the quest to comprehend the processes that regulate climate, why focus on tropical Pacific sea surface temperature (SST)? The equatorial Pacific contains the largest pool of the warmest water on Earth, and as such, is an ideal site to observe how the ocean-atmosphere system might respond to a global warming scenario. It has been observed that sea surface temperatures in this "warm pool" do not rise above 303K (30°C), and that this cap has existed throughout Earth's history. Even 50 million years ago, when dinosaurs roamed the Earth in the Cretaceous period, paleoclimatic evidence suggests that equatorial oceans were limited to the same maximum temperature as now, despite the fact that the climate was much warmer. It is in this context that C4 helped launch CEPEX the most intensive study ever undertaken of this region, from land, sea, air, and space, in an attempt to better understand the processes that act like a thermostat to regulate sea surface temperature.

CEPEX Platform Schematic

The Super Greenhouse Effect and the Thermostat Hypothesis

Scientists have long theorized that warming sea surface temperatures create deep convection in the atmosphere above the ocean and cause that air to become laden with water vapor. Because water vapor traps heat, this process causes additional warming, creating a positive feedback loop that results in a "super greenhouse effect." In CEPEX, this effect was directly measured and proven to exist for the first time. A high resolution radiometer showed that the energy radiated by the atmosphere to the ocean's surface increased so rapidly that the ocean was unable to rid it self of its heat, resulting in a super greenhouse effect. CEPEX also provided observations of processes that limit this effect. For example, deep atmospheric convection causes increased formation of cirrus clouds. These clouds are highly reflective and thus reduce incoming solar radiation, preventing SSTs from rising further. In this way, the cirrus clouds act as a thermostat for SSTs.

C4 investigators Ramanathan and Collins first introduced this cirrus thermostat hypothesis in 1991 to explain two related issues for the present climate: first, why are maximum SSTs so close to the 300 Kelvin threshold for deep convection; and second, what forces limit SSTs and deep convection? Simply put, the thermostat hypothesis states that as SSTs rise above a threshold of 300K, convective storms develop that shade the surface, substantially reducing surface insolation. The reduction of solar radiation in conjunction with other processes cools the ocean and maintains the temperature close to the 300K threshold. CEPEX provided observational evidence of this hypothesis by measuring variations in radiation at the interface of the western Pacific warm pool and the colder eastern Pacific. The investigators also measured cloud microphysical properties in order to better understand the radiative effects of convective cirrus anvil clouds. These measurements, in combination with those collected in the Tropical Ocean Global Atmosphere - Coupled Ocean-Atmosphere Response Experiment (TOGA-COARE), provide an extensive data set on the tropical climate system.

Alternative Hypotheses

Other processes thought to limit SSTs were studied in CEPEX, including the role of evaporation in limiting sea surface temperature. Conventional wisdom prior to the experiment was that a warmer ocean increases evaporation, and that evaporation from the surface helps to lower SSTs. But CEPEX analysis revealed that in fact, there was less evaporation where the ocean was hottest. Surface evaporation did increase with rising SST up to 301K, but then decreased with rising SST at higher temperatures, implying that evaporation is not an effective cooling factor when SST exceeds a threshold value. CEPEX observations also helped advance a theory as to why this is so: there is a decrease in surface wind speed with rising SST in a convectively disturbed atmosphere that may be responsible for the observed decrease in evaporation, as though a cooling fan were somehow switched to a lower speed. This result came as a surprise to many who believed that the evaporation effect was the dominant negative feedback that regulated SST. CEPEX provided the first actual measurements to help settle the theoretical debate.

Visual clues

The Ocean's Heat Budget

CEPEX investigations are centered on the heat budget of an ocean column and its change with SSTs. A comprehensive exploration of thermostat mechanisms necessarily involves a combination of observations, simple empirical models, and coupled ocean-atmosphere general circulation models (GCMs). Such GCMs are able to simulate large time and space ranges that are far beyond the reach of field experiments. The observations from CEPEX and TOGA-COARE are, in turn, useful for testing and validating the physics in these coupled models. These observations come from satellites, aircraft, ships, and the Tropical Atmosphere-Ocean Buoy system (which provides an important, on-going, long -term data base). Because of limited resources and the one-month time period of the field experiment, the CEPEX approach is to "ground truth" the satellite and buoy data with the aircraft and ship data wherever possible, and then use the longer and more widely geographically distributed data for obtaining general conclusions.

CEPEX findings show that convection both heats and cools the ocean surface. Ongoing research will increase our understanding of the net effect of convection on SST (which depends on the interaction between atmospheric forcing and the response of the ocean mixed layer). The CEPEX data show that convection moistens and therefore warms the atmospheric column locally. The convective clouds also heat the surface with longwave radiation, but this forcing is less than 20% of the longwave cloud forcing for the entire column. If the mixed layer is in thermal equilibrium, the reduction in insolation due to cloud shading should be 50% larger than predicted by radiative models. The reduction in solar energy reaching the ground is apparently caused by enhanced shortwave absorption in the cloudy tropical atmosphere. The net effect of convection on the local radiation balance is dominated by this cloud shading effect.

Enhanced Solar Absorption by Clouds

Two notable articles on which C4 scientists collaborated were published in Science, 27 January 1995. The subject of these articles is a key area of inquiry for C 4 and one of the cutting edge issues in current climate research. "Absorption of Solar Radiation by Clouds: Observations Versus Models," by Cess et al ., reports the results of observing the same clouds from above and below to assess how much shortwave solar radiation is actually absorbed by clouds. Co-located satellite and surface measurements of solar radiation at five geographically diverse locations showed that clouds actually absorb about 25 watts per square meter more radiation than predicted by theoretical models. This could result in about 20% less solar energy reaching the ground than was previously believed. "Warm Pool Heat Budget and Shortwave Cloud Forcing: A Missing Physics?," by Ramanathan et al., reports similar findings based on data collected over the tropical western Pacific Ocean during the CEPEX and TOGA-COARE missions. The authors conclude that the shortwave cloud forcing in this region was large, about -100 watts per square meter, and that it exceeded its observed value at the top of the atmosphere by a factor of 1.5.

The results reported in these two articles have dramatic implications for the general circulation models used in climate research. None of these models has accurately reproduced Earth's current climate as yet. But when one of these models (the National Center for Atmospheric Research's Community Climate Model 2, known as NCAR CCM2) was recently altered to make its clouds absorb shortwave radiation in line with the new observations, the results were promising; it compared better with observations of the present climate than the previous version of the model. C 4 investigators will continue to work towards better comprehension of the physics of enhanced absorption by clouds and will use this knowledge to improve the accuracy of the models that may offer us a window on future climate.

Model Evolution

Improving Climate Models

CEPEX data are already helping to modify the general circulation models that attempt to forecast how the planet will respond to increasing levels of greenhouse gases. One such data set is CEPEX's first-ever measurements of a cross section of the water vapor column. This new data is consequential because water vapor is the most significant green house gas several times more powerful than carbon dioxide at directly warming the Earth. The data show that deep convection in the atmosphere (caused by a warming ocean) causes increases in water vapor over a 1000-kilometer scale, literally saturating the atmosphere over a wide area. This new information is not consistent with previous conventional wisdom that convection dries out the atmosphere. It also alerts us that warmer oceans and increasing convection may moisten the atmosphere and enhance the greenhouse effect, much as the models predict.

CEPEX data suggest the following necessary conditions for coupled ocean-atmosphere models to obtain more accurate simulations of the tropical climate:

Climate Model Evaluation

Global Climate Models are the most comprehensive tools for studying Earth's climate system. A climate model includes among other effects the interactions between cloud radiative and convective processes and their effects on large scale dynamics. These models also include the capability for clouds to transport chemical species. Global models must be compared with observational data on various scales to lend credence to the ability to realistically reproduce the present climate. A major focal effort within C 4 is to compare the National Center for Atmospheric Research (NCAR) Community Climate Model (CCM) to field observations. The CCM is a global climate model that is used by a large number of universities for research ranging from paleoclimate to climate predict ability studies. Thus the effort to evaluate and improve this model benefits both C 4 and the wider university community.

The major focus of this effort has been to compare various versions of the CCM with CEPEX data. The evaluation includes comparison of model simulated latent heat flux, boundary layer structure and dynamics, cloud and radiative properties with observations from CEPEX. Initial comparisons between model and data indicated that the model surface winds were too strong and this led to excessive latent heat fluxes, by as much as 80 Wm-2. The next phase of the project was implemented of a new deep convective parameterization developed by G. Zhang at C4. Evaluation of the new CCM including this convection scheme indicated a significant reduction in surface wind speed and an associated reduction of surface latent heat flux. Wind speed over a wide range of the Pacific are now within 1 to 2 ms-1 of observed values. The boundary layer heights are within 50 meters of those observed from ship borne LIDAR. Thus, there has been a major improvement in the climate simulation of the tropical Pacific in the CCM. How ever, improvements extended beyond just the tropical Pacific. The overall dynamical simulation of the model has significantly improved. For example, the winter northern hemisphere stationary wave pattern is much improved over CCM2.

The improvement in the CCM have been so encouraging that this improved version has now become the new standard model, deemed CCM3. Furthermore, this version of the model is being used in the NCAR Climate System Modeling (CSM) project, which has coupled CCM3 to a full depth global ocean model. The next phase of the C 4 Climate Model Evaluation project is to study the effects of enhanced shortwave absorption on the simulated climate. A study by Kiehl et al. (1995) investigated this effect in CCM2 and found significant changes to the simulated climate. A similar study will be carried out in CCM3 to see if the model sensitivity has changed.

Atmospheric Chemistry and Chemical Transport

As part of the effort to enhance our understanding of climate processes, data on atmospheric chemistry, particularly ozone concentrations, were also collected during CEPEX. The data reveal that some of the lowest concentrations of ozone in Earth's atmosphere were found high in the tropical troposphere, just below the tropopause. In pristine areas, ozone concentrations are expected to increase with height, from the surface up into the stratosphere (where other processes complicate matters considerably). The lowest amounts of ozone are expected to occur near the surface, where ozone is rapidly destroyed through chemical reactions. Because these reactions decrease with height, ozone values are expected to grow with height above the surface. The surprise in the CEPEX data was that very low ozone values were found quite high in the troposphere values lower than those seen near the surface.

A recently released paper by C4's Dieter Kley, et al., reports the lowest ozone concentrations ever mea sured in the troposphere below 10 parts per billion by volume over an extended section of the tropical Pacific, an area characterized by intense convection. According to the authors, "The production of close-to-zero ozone mixing ratios in convective regions over tropical warm oceans has several potentially important chemical implications. The very low O3 and NO concentrations and strong reflection of solar radiation from the top of the clouds may lead to low OH concentrations in the convective regions compared to elsewhere in the tropical troposphere. This will result in reduced removal rates of gases that are released into the atmosphere from the tropical oceans. As a consequence, in combination with vigorous upward transport of boundary layer air, a substantial fraction of a number of otherwise relatively short-lived gases released at the Earth's surface may reach the upper troposphere, and maybe even the lower stratosphere. Examples are dimethylsulfide, and organic halide gases with potentially important in fluences on sulfate aerosol formation and lower stratospheric ozone photochemistry."

C4 investigators are currently exploring theories to explain the low ozone phenomenon they observed during CEPEX. Paul Crutzen is leading the exploration of a chemical explanation, meaning that some unknown and unusual chemical process may be causing the very large ozone loss rates observed in the upper troposphere. Dieter Kley is looking into a dynamical explanation, which assumes that the very low ozone amounts seen high in the troposphere result from the transport of low-ozone air from the boundary layer. The final answer may lie in a combination of these two explanations.

Chemical Transport Modeling

These and other questions regarding atmospheric chemistry can be investigated in the context of a chemical transport model. One of C4's major undertakings in this area is called MATCH (Model of Atmospheric Transport and Chemistry). MATCH can use either meteorological fields from GCMs or analyzed datasets to drive the chemical transport, and a variety of convective transport schemes can be used in the model. The model currently includes a set of gas phase reactions and researchers are working on adding the reactions important in modeling aqueous phase chemistry, as well as formulations that will help account for cloud effects.

The ongoing development of the model is being led by C4 investigator Phil Rasch, in collaboration with Paul Crutzen, Mark Lawrence and other colleagues. The team is preparing to further study the low ozone phenomenon using a comprehensive chemistry model developed by Lawrence and Crutzen which has been merged with MATCH. In addition, a parameterization for the representation of cloud water content and cloud microphysics, developed by Rasch for the NCAR CCM2, is also currently being inte grated with MATCH. Results from this merger may be helpful in investigating the theory that the low tropospheric ozone amounts involve chemical reactions taking place in the presence of condensate. The sum of these modeling efforts provides a valuable tool for investigating the low ozone phenomenon as well as other questions regarding atmospheric chemistry and chemical transport.

CIDS: C4's Innovative Data Resource

A unique and exciting outcome of the experiment is the C4 Integrated Data Set (CIDS), a user-friendly data base developed by Erwin Boer, Adam Susman and colleagues, which makes accessible all of CEPEX's raw and derived data sets through a variety of software tools. CIDS makes it simple to access complex information (from micro-scale ice crystal distribution to hundred-kilometer-scale water vapor data to cloud morphology measurements from lidars) in a single system. A computer queried in plain English can co -locate and deliver heterogeneous data from ships, buoys, satellites, and aircraft. This data base is currently being used by modelers to test and validate GCMs, as well as by researchers and students. As part of C4's mission to see that this data is put to the best possible use, it intends to make CIDS widely available, not only to the scientific community, but also to interested students at the high school and college levels. Since March 1995, CIDS is available via the C4's World Wide Webb site. By last count in early November, the number of accesses was fast approaching the 10,000 mark with about 18,000 files transfered.


Return to Publication page

Last Updated : 11/24/97