This NSF Science and Technology Center (STC) will focus on the representation of cloud processes in climate models. The STC's name is the "Center for Multi-Scale Modeling of Atmospheric Processes" (MMAP), and the lead institution is Colorado State University (CSU). The goal of MMAP is to break the "deadlock" that has stalled the progress of climate research for several decades. Climate models are physically based and include representations of the atmosphere, the ocean, the land-surface, and the cryosphere. They run on the most powerful computers available. They are now providing predictions of future climate change due to anthropogenic changes in the composition of the Earth's atmosphere. These predictions are being used as input to policy decisions that have enormous economic implications for the U.S. and the world. It has been true for decades now that our inability to simulate the interactions of clouds with large-scale atmospheric circulations is one of the most important limitations on the reliability of climate-change simulations. Poor simulations of cloud systems also reduce the skill of weather forecasts, especially for precipitation. MMAP will address this problem through a revolutionary new approach called the "multi-scale modeling framework" (MMF), in which fine-grid Cloud-System Resolving Models (CSRMs) are embedded within the much larger grid cells of an atmospheric general circulation model (GCM). In an MMF, the CSRM takes the place of the single-column "conventional parameterizations" that are used in current GCMs. Whereas conventional parameterizations are based on statistical theories involving uncertain closure assumptions and little or no information about the spatial structure of the cloud field, MMFs resolve cloud processes explicitly down to a scale of a few kilometers, and so represent some aspects of the spatial structure explicitly.
The first MMF was created by MMAP scientist W. Grabowski of NCAR. In most of the prototype studies carried out to date, the CSRM is two-dimensional (2D), with periodic boundary conditions. It represents a "sample" of the clouds in a GCM grid column. The CSRM's high-resolution depiction of a cloud field can be used to compute statistics (e.g., the precipitation rate and fractional cloudiness) for the sampled portion of the GCM's grid column, and these statistics are applied to the entire grid column. A key point is that MMFs can represent the cloud-scale interactions among the many physical and chemical processes that are active in cloud systems, including cloud dynamics, microphysics including aerosols, turbulence, and radiation. MMFs eliminate the need for closure assumptions to determine the strength of deep convective activity. They eliminate the need for cloud-overlap assumptions in the radiative transfer and microphysics parameterizations. They have the potential to represent the interactions of both clouds and gravity waves with orography. They are also particularly attractive for the simulation of chemical species transports and transformations within cloud systems, as well as small-scale interactions between the atmosphere and the biological and hydrological processes of the land-surface. MMFs must still include parameterizations of critical sub-cloud-scale processes, including microphysics, turbulence, and radiative transfer. Because these processes are represented on the cloud-scale, however, they can be parameterized in relatively straightforward ways. A further very important strength of an MMF is that the results produced can be evaluated by comparison of simulated and observed cloud-scale processes. Recent work at CSU has shown that, relative to a control simulation with a conventional GCM, a prototype MMF produces greatly improved simulations of atmospheric variability on a variety of time scales, from diurnal to intra-seasonal. It also gives more realistic simulations of cloudiness and precipitation. Experiments with the MMF have already shown that cloud-scale variability of the radiative heating rate is important, as is convective momentum transport, which is included in a new version of the MMF that uses a 3D CSRM. A key part of the research consists of further development of the MMF concept, beginning with a new version of the MMF in which the periodic boundary conditions of the CSRM are eliminated, and multiple 2D CSRMs are combined to create a "quasi-3D" MMF. Removal of the 2D constraint permits convective systems to have arbitrary orientation and to vertically transport horizontal momentum. Removal of the periodic boundary conditions allows convective systems to propagate from one GCM grid column to the next, and prevents the convection from being artificially "squeezed" as the periodic domain decreases in size. Because there is no reason to alter the formulation of the embedded CSRM when the GCM's resolution is increased, the formulation of the quasi-3D MMF is independent of the spacing of the outer grid. In addition, realistic topographic forcing can be prescribed from data, and used to simulate orographic gravity waves and orographic clouds. Finally, a quasi-3D MMF converges in a smooth and natural way to a global CSRM, as the size of the outer grid is refined. The PIs emphasize that no existing GCM has this convergence property. They plan to develop, evaluate, and apply a quasi-3D MMF as the central, organizing activity of the research.
MMAP's education and human-resource goals are to provide first-rate graduate education in Atmospheric Science; to interest undergraduates in graduate education and careers in climate science; and to develop and disseminate teaching materials designed to inform K-12 students (and their teachers) about the nature of the climate system and the career opportunities in climate science. In each of these areas, MMAP will make a special effort to include students from groups that are under-represented among climate- science professionals. MMAP will undertake two publishing projects that will significantly enhance scientific communication in our field: the creation of a new and unique online technical journal devoted to global modeling, and the production of an edited book on the history of global climate modeling, including transcripts of interviews with the key participants.
The intellectual merit of MMAP's research lies in its revolutionary approach to the cloud-climate problem. Earth system scientists can learn a lot about the global climate system by approaching the problem of climate modeling from a new and different perspective, and this new knowledge is the most valuable thing that will flow from MMAP's research.
The research will have broad impacts on both science and society because it will increase both our understanding of climate dynamics, and our ability to make reliable predictions of cloud feedbacks on climate change. The legacy of MMAP will include important new modeling tools that will provide substantially more reliable predictions of anthropogenic climate change. In addition, MMAP will demonstrate new ways to compare high resolution observations with global model results, enable improved weather forecasts by the operational centers, strengthen the scientific interactions between global modelers on the one hand and cloud-scale observers and cloud modelers on the other, create a heightened awareness of the excitement and opportunities of climate research among both female and male students from all ethnic backgrounds and at all levels, inaugurate a unique new scholarly journal, and produce a book that captures the history of global climate modeling with an emphasis on the cloud parameterization problem.
