FIRE I Strategy

The FIRE-I Implementation Plan (1985) outlines a series of investigations and observations designed to meet the goals of basic understanding and parameterization. In this plan, FIRE-I is meant to address the issues of basic understanding and parameterization of cirrus and marine stratocumulus cloud fields and ISCCP data products.

Besides sharing a common set of goals, the three FIRE-I components share a general set of strategies adopted to achieve those goals. These three common strategies are briefly set out below:

Strategy 1:

INTENSIVE FIELD OBSERVATIONS - Intensive field programs are currently planned to support all three components of FIRE-I. Two field programs of three to six weeks duration are planned. These field programs are designed to observe cloud systems with high resolution temporal and spatial sampling. The field programs will include satellite, aircraft, balloon borne and surface observations. The experiments will be concentrated in limited geographical areas for periods of three to six weeks. Two intensive field programs are planned. The Cirrus Intensive Field Observations (IFO) will be conducted in Fall 1986 in the continental U.S. with an emphasis on the study of cirrus systems. The Marine Stratocumulus IFO will be conducted in summer 1987 near San Nicolas Island, off the west coast of California and will emphasize marine stratocumulus systems. Although the intensive field operations will emphasize the cirrus and marine stratocumulus objectives, they would also strongly contribute to the ISCCP intercomparison objective.

Strategy 2:

EXTENDED TIME OBSERVATIONS - We have much to learn from data bases already collected and those currently being collected; this is particularly true for satellite data and to a lesser extent for other types of data. The extended time observations (ETO) of cloud systems will yield information on the phenomenology of these systems, on their variability in space and time, and on their bulk radiative properties. Since these data will generally be of higher spatial resolution than the ISCCP data set, they will be particularly useful for comparison with ISCCP products.

Strategy 3:

MODELING - The modeling strategies for FIRE-I encompass several types of models. These models range from radiative transfer models which require cloud microphysical data as input, to cirrus and marine stratocumulus models capable of describing the time evolution of these cloud systems, to general circulation and climate models, which will ultimately utilize cloud parameterizations, and to the retrieval algorithm models utilized in ISCCP. All FIRE-I modeling strategies seek first to provide a means to compare our best current understanding of a phenomenon with observations of that phenomenon and second to provide a means to extend that understanding by utilizing the models to extrapolate to other conditions.


    Intensive Field Observations (IFO) programs are planned to support research tasks requiring high time and space resolution information on cirrus and marine stratocumulus cloud systems. In addition to supporting the cirrus and marine stratocumulus studies, these data will be instrumental in the development of a better understanding of ISCCP data products. The IFOs will be gathered from a variety of platforms on a relatively local, but regionally representative, geographical scale. Data will be collected from multiple satellites, aircraft, balloon and surface-based instrumentation. The intensive data collection periods will be approximately three to six weeks in length and will be scheduled annually starting in the fall of 1986.

    The sampling strategy involves obtaining observations of the same targets from multiple platforms. Synchronized high resolution multispectral satellite observations will be obtained from multiple platforms viewing the scene from different angles. Coincident observations will also be obtained from aircraft where one is essentially a satellite platform simulator with additional capabilities and the other aircraft are equipped with in situ radiative, microphysical, and air motion sensing instrumentation. Special lidar, radar and radiometric observations will be obtained from surface based sites as will conventional observations of meteorological parameters. The strategy calls for high density observations over limited times (2-3 hours) both in small (10 km)2 fixed regions where surface based observations are being collected and over surrounding regions as weather and satellite operations permit. Observations over the fixed region will serve the purposes of the intensive case studies while missions over the surrounding regions, particularly over water, will provide important data for cloud retrieval algorithm intercomparison.

    Two intensive field programs are planned. The Cirrus IFO will be conducted in Fall 1986 in the continental U.S. with an emphasis on the study of cirrus systems. The Marine Stratocumulus IFO will be conducted in summer 1987 near San Nicolas Island, off the west coast of California and will emphasize marine stratocumulus systems.


    The Extended Time Observation (ETO) program will directly support the ISCCP and GCM validation efforts. The ETO data will consist of satellite observations, special ground stations with lidar and surface radiation measurements, and conventional meteorological observations. These data will provide a means of extending the results derived in the more detailed Intensive Field Observation Program intercomparison studies to larger time and space scales.

    The Extended Time Observations are subdivided into two space scales: Extended area and Limited area (Figure 1). The extended area data set is meant to provide data over a large geographical area where occurrences of cirrus and stratocumulus cloud systems may be found in a variety of geographical locations; and to allow for multi-satellite, multiple-view observations of these systems. The limited area data set is geographically specific to the location and surrounding area of surface observing sites being maintained throughout the FIRE-I experiment.


    An obstacle to reliable predictions from first principles of clouds and their radiative effects in GCMs is the lack of understanding of the large scale statistics of turbulent atmospheric motions and the clouds formed by these motions. The ISCCP research strategy to overcome this obstacle is to coordinate a number of intensive, smaller scale projects with a global climatology effort. The goal of FIRE is to improve GCM cloud-radiation parameterizations, for at least cirrus and marine stratocumulus clouds, by providing a link between the small scale cloud processes and the planetary scale climate processes. This goal is to be accomplished by the interaction of data analysis and modeling studies on a range of space/time scales. Progress comes through the iterative application of four steps:

    1. Improve cloud radiation models.
    2. Improve cloud dynamics models.
    3. Intercompare the ISCCP cloud climatology with FIRE cloud data.
    4. Conduct comparisons of GCMs with the improved cloud-radiation, dynamics models and with the FIRE and ISCCP observations.

Each of these steps is to be carried out for marine stratocumulus and cirrus clouds as part of two types of studies: (1) The intensive field observations (IFO) programs will collect and analyze limited area/time data focused on the diagnosis of cloud processes at smaller scales; (2) Extended time observations (ETO) will collect and analyze regional, multi-year data to bridge the gap between IFO and ISCCP data. The unique attribute of the FIRE IFOs is the coordination of satellite, aircraft and ground observations. This section outlines the general modeling tasks planned as part of FIRE.


(Step 1 and 4)

This task addresses the problem of calculating the radiative effects of clouds given a specification of cloud properties. The primary issues are to determine which properties are most crucial to accurate calculations of radiation and to ascertain the sensitivity of radiative calculations to cloud parameters that are difficult to measure. Since most cloud radiation models involve approximations, for example ignoring small scale variations in liquid water content, this task is also concerned with testing such approximations. The strategy for FIRE is to assemble data sets which contain both specifications of cloud and atmospheric properties and independent verification measurements of the radiation field produced by the clouds. The former are needed to define the clouds in the radiative transfer model (input) and the latter are used to verify the calculated radiation (output). The observations must be balanced between input and output parameters to provide definitive tests of model performance. These data will be used to test various types of radiative models, particularly column models and full three-dimensional models. Intercomparisons of the models provides further insight for parameterization of cloud-radiation effects In GCMs.


(Steps 1 and 3)

This task addresses the problem of inferring cloud properties from measured radiances that is central to the analysis of FIRE observations and constitutes a major contribution to the interpretation of the ISCCP cloud climatology. Improvement of radiative models that relate cloud properties to observed radiances is not only intimately linked to the general understanding of the cloud-radiation interaction in Steps 1 and 4, but also includes specific problems raised by the analysis of satellite data in ISCCP. Satellite-measured radiances are generally limited in spatial and temporal resolution, angular coverage, and spectral coverage; thus, retrieved cloud parameters using current methods depend on these factors (Rossow et al., 1985). These factors also present a challenge to the coordination of satellite data with other observations for retrieval validation. The IFO observations must be coordinated to provide a balanced set of input radiances and output cloud properties to test retrieval techniques. Limited intercomparisons using ground-based and aircraft observations must also be extended to larger scales by statistical comparisons of multi-satellite data. Comparisons of retrievals by different techniques applied to the same FIRE and ISCCP data are a key to development of better methods.


(Steps 2 and 4)

This task concerns the understanding of the processes that produce clouds from particular atmospheric states. The special emphasis of FIRE is to extend our understanding of the workings of these processes on the smaller space/time scales of individual clouds to cloud structures and the state of the atmosphere on meso- and synoptic scales. The primary data for this task will be that obtained during the IFOs targeted on marine stratocumulus and cirrus; however, the ETO, based largely on satellite data, is necessary to link small scales to large scales. This task also depends on radiative model improvements to increase the detail retrieved form the IFO and ETO data. The emphasis is on model comparisons to data to diagnose processes and to improve modeling of these two cloud types.

Cirrus clouds exhibit significant horizontal and vertical structure and a strong coupling to radiation (Starr and Cox, 1985). Little is known about the distribution of cirrus, the conditions required to produce them, or their life cycles. The primary focus of this task is to improve understanding of the coupling of cirrus clouds, radiation, and atmospheric dynamics by comparisons of models to coordinated measurements of all of these quantities. Models that will be employed for these studies include two-dimensional (horizontal/vertical) time dependent process models, high-resolution, limited-area models and GCMs. The coordination of ground, aircraft and satellite observations during the IFOs, as well as the extension of the IFO results to the ETO and ISCCP scales, is crucial to this task.

Marine stratocumulus clouds exhibit considerable structure from the smallest scales (-1 km) resolved by satellites (Coakley and Bretherton, 1982) to mesoscale/synoptic scale (Agee et al., 1973). However, only a few modeling and in situ observational studies have focused on fractional cloudiness (Sommeria and Deardorff, 1977; Albrecht, 1981; Bougeault, 1982). A major focus of this task is to elucidate the processes responsible for determining fractional cloud cover through comparisons of models to FIRE data. A broad spectrum of dynamical models will be employed, including simple mixed-layer models, "large" eddy models that explicitly resolve the primary turbulent motion scales, high resolution, limited area dynamical models and GCMs that include boundary layer and cloud parameterizations. Microphysical models will also be used to investigate the coupling between dynamics and radiation through cloud radiative properties.


(Step 4)

This task includes many of the activities discussed above but the emphasis is on the climate GCM calculation of radiative flux divergences from a particular large scale atmospheric state involving clouds. The FIRE concept breaks this calculation into two parts: calculation of cloud properties from an atmospheric state and calculation of radiation fields from cloud properties. The primary focus of this task within FIRE is to define quantitative methods for comparison of data and models; this is an appropriate FIRE task because of the detailed study within FIRE to understand the linkage between IFO and data and between these data and specific process models. The FIRE data sets allow for GCM tests in both prognostic and diagnostic modes. FIRE and ISCCP data, together, provide a check of the statistical behavior of GCMs, while FIRE and ERBE provide a check on the statistics of the GCM radiation budget. Comparisons of data to GCMs must build on the FIRE developments that link multi-scale observations and model results and develop methods for comparison of cloud properties, radiation quantities, and the space/time statistics of these.