GEWEX Cloud System Study: Working Group 3
Data Blueprint
TABLE OF CONTENTS3. INTERNAL THERMODYNAMIC, MOISTURE AND DYNAMIC FIELDS
5. RADIATION
5.1 In-Situ measurements
5.2 Satellite Measurements
9. SCHEMATIC DEPICTIONS OF THE WEATHER SYSTEMS
10. AIRCRAFT DATA CONSIDERATIONS
1. INTRODUCTION
For the goals of WG3 to be satisfied, appropriate data must be acquired to address the scientific issues raised in the extra-tropical cyclone section of the GCSS Science Plan. The data are needed for initializing and validating cloud and mesoscale models so that they can be used as reliable tools for improving parameterizations of these weather systems within climate models.
The key scientific issues include: heat and moisture budgets, internal evolving fields including slantwise ascent, cloud layering, radiative attributes, microphysical interactions, and precipitation. A range of observations are needed to properly address these scientific issues. Large scale, fairly coarse resolution information can be observed by arrays of soundings and satellite data but within this meshing finer scale observations need to be made with earth-based remote sensing and in-situ observations.
Historically, experiments on extra-tropical cyclones have suffered from a lack of appropriate information for large scale characterization in general and for model initialization in particular. In terms of system evolution, perhaps one of the main uncertainties not well-sampled is the genesis stage of a cyclone from an initial flat flow. Fine scale measurements have been carried out to study processes in specific regions of storms, but the large scale settings within which these can be placed have usually been inadequately measured. Without the necessary wide distribution of measurements, the heat, momentum and moisture budgets of systems have not been well enough known to validate the overall outputs of large scale models and their parameterizations. From the point of view of process studies, the lack of mesoscale data over adequate regions (particularly vertical velocity and momentum fluxes) has long been a limitation.
There is then a need for a wide range of measurements within the context of GCSS. There are two principal motivations for this. First, they are needed to provide information to validate models and, second, to provide information to improve them through process studies. The organization of the following sections corresponds roughly to a progression from validation to process description with descending scales.
The capabilities of measurement systems are a strong constraint for almost all of the desired measurements. Very careful pre-planning of activities and the close development of plans between different instrumentation specialists will be required. The first need in this regard is an assessment of model performance by modellers and a statement by the modellers of the sensitive issues. It may well be that some critical fields are predicted quite well already. The experiment should then concentrate on critical aspects not well predicted.
Table 1 and Sections 2-8 present many of the parameters that need to be observed, as well as their expected values and scales, to satisfy the data requirements of each of the main issues identified in the Science Plan. The text also summarizes some of the considerations that were made in arriving at these estimates. Section 9 schematically illustrates some of the regions where particular features are believed to occur and Section 10 summarizes some of the aircraft flight considerations involved in obtaining the necessary information.
2. CLOUD SYSTEM BUDGETS
Because of the low resolution of large scale models, a first need is validation of the overall budget and the transfers by which this is achieved. Also, the large scale environment needs to be properly measured in order to initialize cloud resolving models and to determine moisture, heat and momentum fluxes.
It is expected that these features can be obtained from special rawinsonde and/or dropsonde measurements over a suitably large region surrounding the storm itself. Spatial resolution of about 100 km along tracks of at least 1000 km is needed. Such measurements are obtainable with available dropsonde technology; the impending use of the Global Positioning System for windfinding is essential to improving the accuracy of the soundings. These measurements will supplement the fields generated from NWP models. The proper use of limited resources to adequately accomplish these supplementary observations will however require detailed planning.
Observations near regions of tropopause folds will probably need to be finer resolution. The scales of these folds are such that measurements down to about 10 km may be necessary to properly quantify the exchanges between the stratosphere and troposphere. Aircraft measurements are probably the best way to ensure such observations.
The high variability of water vapour provides particular problems in evaluating the budget. Low level fluxes vary greatly, whereas the smaller upper level fluxes can be contained within narrow regions, particularly around jet streams. However, use can be made of satellite information to examine these smaller regions that are not resolved by the sounding program.
The measurement of moisture is also difficult. For example, many standard sondes do not work well below about 20% relative humidity and so one is faced with examining aircraft-based measurements of high accuracy within a few critical sectors of the overall cloud system. Satellite-based measurements of the needed vertical resolution produce averages of the order of 50 km across; the techniques of using this information have furthermore only been applied to low levels.
The nature of surface conditions can play a major role in determining the characteristics of the cloud systems. With available satellite-based systems for inferring surface parameters, 5 km information is largely obtainable in cloud-free conditions for sea surface temperature and sea ice fractions. The actual surface fluxes can be deduced from several present satellite systems, but it is likely that there will be a need for in-situ aircraft, ground and ship based measurements as well.
3. INTERNAL THERMODYNAMIC, MOISTURE AND DYNAMIC FIELDS
The overall budget for a system depends upon its internal dynamical processes and their interactions with the ambient larger scale flow, a principle which operates through a descending continuum of scales. Thus, the response to imposed forcing and the development of baroclinically and otherwise unstable structures are essential to explaining the budgets. Moisture is not passive in this system, since it provides major components of the forcing through latent heat transfers. These internal features are briefly discussed in the following sections.
Several forms of instability are believed to occur within the different regions of the storms. These need to be assessed and they each have associated with them different criteria and features that need to be adequately sampled. For example, within the cloud field associated with the warm front, the internal variability can be a result of local convective instability, symmetric instability, mechanically and diabatically forced circulations, and inertia-gravity waves. To assess convective instability of horizontal scales of the order of 5 km, then sampling every 1-2 km is needed to avoid aliasing. To assess other circulations on scales of the order of 50 km, then sampling every 15 km is needed. Many of these measurements can only be carried out with aircraft, although remote sensing tools such as Doppler radar are essential for adequate sampling.
It is important, but difficult, to observe the vertical transport of heat, momentum and moisture in frontal systems. Organized mesoscale ascent occurs on scales from a few hundred meters to about 100 km. The narrowest regions are furthermore generally linked with non-hydrostatic forcing. Locations near frontal surfaces, regions of convective activity, as well as transport near the tropopause are a few of the areas needing particular attention. Doppler radar and aircraft observations will be crucial to the proper measurement of these parameters. Near the tropopause in particular, Doppler lidar measurements should be very important.
4. CLOUD LAYERING
Thin layer clouds related to frontal systems can substantially alter the radiational properties of the systems, and in some regions of the storms there can be multiple cloud layers. These layer clouds vary from about 100 m to over 1 km in thickness. Their presence and composition as well as their horizontal variations need to be assessed. Cloud sensing radar will lead to a much better knowledge of layering within these systems, although satellite-borne instruments will be needed to achieve a large scale perspective that is not attainable by with these ground-based or aircraft based instruments.
The tops and bases of clouds are sensitive factors for understanding the distribution of atmospheric radiative heating and the surface radiation budget which are needed for climate studies. In addition, it appears that NWP models often predict cloud tops which are higher than observed; this will introduce a bias into radiative calculations.
The main limitation of current satellite instruments for deducing clouds is that they mainly sense cloud top. Present systems do no better than about 300-500 m in vertical resolution. This reflects the uncertainties in, for example, infrared retrieval techniques which are only accurate to about 1.5K. However, plans for the proposed GEWEX Cloud Profiling radar has rationalized that 500 m is reasonable since this is sufficient to give 10 w m-2 accuracy.
Processes occurring close to cloud base and top (affecting radiational, dynamic and physical properties of these regions) are nevertheless very important. Detailed aircraft-based measurements are consequently needed to examine this aspect of the clouds so that they can be used to tune satellite measurements and be properly represented within all-encompassing models.
In addition, it is certainly critical to know when and where multi-layering occurs and what the characteristics of these layers are. It may be that in some instances, the large scale impact of such layering is minimal; it is the base of the lower one and the top of the higher one that are critical. The exact situation needs to be sampled with cloud profiling radars, as well as some in-situ observation.
Over the cloud field of a complete extra-tropical cyclone, the cloud fraction will furthermore vary from complete overcast to scattered cloud. The exact nature of this distribution within the overall context of the storm needs to be well assessed.
Cloud morphology changes dramatically but in a well-organized fashion across the expanse of the systems. Variations in the cloud scape at the top of the clouds show this, and such variations are linked with alterations in radiation, momentum, etc. budgets. Furthermore, the proper sampling of this feature can be used to place more detailed measurements of fine-scale quantities into proper context. Cloud top features also vary horizontally on the 500 m scale proposed.
It should be added that recent observations suggest that ice crystals can exist above the satellite-derived cloud top and these may contribute substantially to the moisture budget of the storms, as well as to the mis-determination of cloud top from satellite by distances of the order of 1 km. If these suggestions are correct, such crystals may need to be measured within the well-defined larger scale environment.
5. RADIATION
5.1 In-Situ measurements
A major aspect of these weather systems is their long and short wave radiational absorption and scattering fields. These must be well observed in order to act as a strong verification data set to cloud resolving models. These measurements can be made from a combination of remote sensing platforms, high-flying aircraft, in-situ aircraft, and ground-based sensors.
For this GCSS initiative, it is felt that the most critical measurements for radiation will be fluxes as opposed to radiances. However, vertical radiance measurements would provide useful information on horizontal inhomogeneities in the clouds.
To properly assess the radiational properties of the systems, it is important to measure both broad and narrow band information in the short and long wavelengths. Satellites generally only utilize narrow bands; broad band information is needed from aircraft observations to determine how the two measurements are related. Narrow band measurements on the aircraft, if feasible, should be chosen to duplicate at least some of the channels being used by the satellites. It is then critical that as many of the aircraft flights as possible be made within satellite coverage.
The emphasis in the acquisition of radiation data should be placed on information from platforms situated below and above the clouds as opposed to within the clouds. It is felt that if in-cloud microphysical information is available (see following Section), then there should typically be enough information to predict the effect of the cloud structure on the out-of-cloud measurements. If some pure ice or mixed phase clouds cannot be adequately handled by this technique, then more emphasis will be placed on inside cloud measurements in those areas. It needs to be established though that the in-cloud measurements could be made with sufficient accuracy to make a difference.
It is important to measure the radiational pattern below cloud base. In experiments over the ocean in particular, such measurements may have to suffice as measures of surface radiation.
Through the examination of satellite information, it is apparent that there are patterns in the radiational structure of the systems. Different regions are characterized by different scales of variability. For example, the deep cloud region linked with the warm front appears to exhibit substantial variability in the short and long wave radiation fields on scales of the order of 100 km, whereas variability in the cold air outbreak region will be on scales less than a few kilometers. Aircraft flights will need to be made over at least a few of these scales of variability.
The overall strategy, including which wavelengths to emphasize within which particular cloud regions, is a critical issue to be resolved. Key input to this strategy will be an assessment of the critical uncertainties that can be addressed with the available instrumentation.
5.2 Satellite Measurements
Satellite-based radiation measurements are already one of the strongest basic verification datasets for climate models. The new generation of geostationary satellites will furthermore have a complement of radiometers that closely resembles that available now on the polar orbiting ones. It is clear that many of the aircraft projects will need to be well coordinated with the times of passage of these satellites.
The ability of geostationary satellites to provide continuous sampling of the infrared and shortwave radiation fields associated with the storms is absolutely essential. The weather systems evolve and detailed in-situ measurements can only be made within short periods of the overall evolution. The geostationary data places these in their proper context.
Information from high resolution polar orbiting and new generation geostationary satellites are needed to examine the fine-scale radiation perturbations occurring within the overall cloud field and to assess the moisture fields associated with weather systems. In the absence of substantial cloud, TOVS information will furthermore produce estimates of the vertical temperature profiles.
Information from the DMSP satellites is also necessary. Through the use of several frequencies of radiation and its polarization, it is possible to infer important characteristics related to the moisture fields including liquid water path, rainfall rate and, possibly, the occurrence of ice particles.
Techniques of deducing such variables have not all been validated however; within the highly variable cloud environment of an extra-tropical storm, it may well be that a great deal of caution has to be applied to some of the results. Hence, the importance of combining in situ measurements with simultaneous satellite data. As one example, it has recently been shown that microwave retrievals of liquid water within deep stratiform regions of the systems can be in error due to the influence of wet snow within the melting layer. With the proper in-situ measurements of vertical profiles in such instances, it may be possible to deduce the proper results although this has yet to be demonstrated.
6. MICROPHYSICS
Sensitivity is the first issue to be addressed in connection with the measurement of microphysical properties. What level of detail is necessary in models depends upon the intent of those models. Issues of radiation, precipitation, mesoscale circulations, static stability alteration, and each overall budget will place different demands on the level of observational detail that is necessary.
Microphysical properties and processes vary across and through the overall cloud fields. Most of these properties can only be observed with in-situ aircraft. Such measurements can be made in fine detail but one has to compromise on the coverage area. Carefully planned flights made in collaboration with satellites as well as other aircraft with lidars and Doppler radars are needed to place such measurements into a larger scale context.
Within an extra-tropical cyclone, the phases and forms of cloud and precipitation particles vary enormously. Different regions of the overall cloud field with their different characteristics may then have widely varying radiation consequences. Furthermore, the actual spectra and shape of the particles can directly alter the radiational properties. Within some regions of the storms such as near fronts and within embedded convection, the spatial variation of these quantities is quite high and will necessitate observations at the finest scales indicated in Table 1.
Some regions of the clouds contain mixed phase particles (those containing both liquid and solid forms of water), some regions contain particles of different phases, whereas others are comprised of particles of only a single phase. Largely because of their radiational importance but also because of their critical relation to, for example, precipitation production, such regions need to be carefully identified. Many processes govern the balance of phases in these situations, for example ice crystal multiplication and ice nucleation. The means through which such factors can be handled will depend upon the resolution of the models. Low resolution models dealing with an aggregate of many cloud regions cannot handle these factors directly.
Within the widespread cloud region, there are characteristic levels at which significant cloud water may be present, though its occurrence within systems is likely linked closely with mesoscale vertical motion that has in general not been adequately observed. These levels are, for example, cloud top, embedded convection, frontal surfaces, near the melting layer, and low level stratus cloud. The depth of the liquid water region near cloud top is normally quite shallow (about 100 m), shallow regions also occur in connection with the melting layer and low level stratus (< 1 km). Typical values of the expected water content and estimates of the needed vertical resolution for observing these are shown.
Within convection or near some of the frontal surfaces, cloud droplet spectra vary enormously, on horizontal scales such as shown. However, it is not clear whether these variations need to be well-sampled; it may be that the microphysical properties of these relatively small regions of the storms provide an insignificant impact on at least the overall radiational properties.
The nature of aerosol particles can lead to significant impacts on several aspects of the storms of importance to climate. For example, the droplet spectra within the associated stratocumulus cloud will be closely linked with the aerosol characteristics. In addition, the onset of glaciation is strongly affected by the ice nucleating abilities of some of the aerosol. Within deep cloud fields, the aerosol particles will not be as crucial to defining the cloud properties; however, many of these systems produce low to mid layer clouds which are particularly sensitive to the aerosol properties.
7. PRECIPITATION
Precipitation at the surface is one of the most important consequences of these cloud systems. It occurs over domains much smaller than the overall cloud field and it varies
substantially on scales of the order of 1 km up to 100s of km. The effective measurement of this parameter can best be accomplished with a combination of surface direct observations as well as remote sensing such as with radar.
In some regions or seasons, the form of precipitation is also variable; it can in general be either rain or snow. Through remote sensing by, for example, polarization radar, it is generally possible to infer precipitation type.
It should be added that many winter storms contain regions of both rain and snow, as well as freezing precipitation. The heaviest precipitation rates within such storms often occurs within the approximately 100 km wide transition region separating these types of precipitation. Such transition regions can furthermore extend for about 1000 km parallel to the warm frontal region in particular. Properly incorporating precipitation related to this feature will need to account for embedded circulations giving rise to the enhanced precipitation, but radar observations to infer precipitation rates will be complicated by the varying forms of precipitation.
8. TEMPORAL SAMPLING
A definitive data set will require the establishment of the temporal evolution of the system. The consistency of two or more sets of observations over a period will furthermore act as an important test of budget estimates.
A period of 6-12 h is probably the best that can be hoped for in a complete re-examination of key components of the system using in-situ observations. Continuous monitoring of the system by means of remote sensing will nevertheless be crucial for developing an appreciation of the optimal re-examination period and, in later analyses, for placing these detailed but dispersed finer measurements into a proper perspective.
9. SCHEMATIC DEPICTIONS OF THE WEATHER SYSTEMS
The portions of the overall cloud field that are linked with the key questions of interest to GCSS must be carefully considered. This rationalization process is required to best utilize the resources available in any field experiment.
There are two categories of measurements that are needed:
- background measurements to improve fields from NWP which can be used to initialize cloud resolving models. These measurements are required for any experiment designed to address the heart of the GCSS initiative.
- specific measurements to examine specific aspects in more detail. The nature of these depend upon the focus of the study.
Figures 1-3 show simple schematics of an extra-tropical cyclone in order to assist in placing these considerations into a better perspective. This organization, with very little weather along the surface cold front for example, is quite common along the eastern Canadian coast in winter but other regions are characterized by other types of organization in which, for example, the cold front is much more pronounced. Figure 1 shows a simple frontal organization onto which typical conveyor belt flows are shown. Regardless of the particular organization, the mapping of the conveyor belt flows is critical to budget studies. Figure 2 emphasizes the cloud fields that can be linked with such systems. Indications of cloud layering, cloud morphology and precipitation are also shown. Figure 3 highlights some of the dynamical, microphysical and other processes that can be occurring within different regions of the overall cloud field.
While these are only schematics and real systems involve several fold more complexity, it is apparent that the high resolution data requirements of some of the parameters do not need to be made over the whole region of a weather system. An important practical topic for research is to apply theoretical tools and diagnostic studies to indicate which factors determine the locations of the important features in particular systems.
10. AIRCRAFT DATA CONSIDERATIONS
Individual aircraft can only sample small portions of the overall cloud field associated with an extra-tropical cyclone. The use of these facilities must then be very carefully considered before-hand as to how to optimize their use.
The assessment of optimum use in a given situation will need to take advantage of the theoretical framework being established, as discussed above in Section 9. This means that it is also essential to rank the critical issues that must be addressed before an actual new field experiment. Everything can not be done simultaneously.
Based upon the intent of the actual aircraft missions, there are a number of flight plans that need to be developed. These fall into two broad categories as summarized below.
Large-Scale System Mapping:
- Model initialization and budget studies
- long high-altitude legs with soundings upstream of the system
- high-altitude legs with dropsondes focussing on the northern and eastern sections for mapping transports from the system
- legs with dropsondes on the southern and eastern regions in particular to map out moisture and momentum transport into the system. Low level information would need to be achievable with the dropsondes or else special low level passes with lidars and possibly radiometers would be needed.
Small-Scale Studies:
- Microphysical properties
- spiral descent. Radius of about 5 km or less to document the vertical structure of the properties.
- constant altitude flight legs. These need to be conducted at several locations within the overall cloud field, including cloud top and cloud base.
- concurrent with radar and satellite measurements.
- Radiative properties:
- similar to the above flight plans
- concurrent with cloud sensing radar observations
- emphasis below and above cloud
- Dynamical properties:
- constant altitude passes and taken along and across frontal zones in particular.
- Large scale cloud effects:
- long constant altitude passes within sectors of the clouds felt to be critical to the large scale impact of the weather systems.
- microphysical data in collaboration with satellite information
11. CONCLUDING REMARKS
This document is considered a necessary first step towards planning field experiments to validate and improve parameterizations of cloud systems associated with extra-tropical cyclones. It should be apparent from the scope of this document that the proper parameterization of this cloud system can only be developed in a systematic manner if a large suite of complementary measurements are made. Such an effort will undoubtedly require international collaboration to reach a critical threshold.
The actual planning of the observational program of such an experiment will of course be done in much greater detail once the availability and logistical constraints of various platforms are considered.
January 4, 1995
Table 1: DATA REQUIREMENTS for EXTRA-TROPICAL LAYER CLOUD STUDIES
Parameter Horizontal Vertical Absolute Expected
Resolution Resolution Accuracy Values
WATER AND MOMENTUM BUDGETS AND LARGE SCALE ENVIRONMENT:
Large Scale Conditions
-sondes 100 km 100 m 1 m/s 100/10 kPa
-water vapour 100 km 100 m 10% >0.5 ppmv
-surface pressure 100 km - 0.1 kPa 101/93 kPa
Tropopause-Stratosphere Region
-dynamics/thermo. 10 km 100 m
-ozone 10 km 100 m
SURFACE CONDITIONS:
Sea Surface Temp. 5 km 0.5oC -4/25oC
Sea Ice Fraction 5 km 10% 0/1
Topography & character5 km
Surface Fluxes 5 km 100 m 10% -100/2000 W/m2
-heat,moisture, mom'm
INTERNAL CRITICAL FIELDS:
Instabilities Temp/winds/humidity
-upright 2 km 200 m " " "
-slantwise 15 km 200 m " " "
-gravity waves 1 km 200 m
-turbulence
Updrafts/downdrafts
-mesoscale 5 km 500 m 10% -5/5 m/s
-small scales 5 m 200 m 10% -10/10 m/s
CLOUD FIELDS
Multiple Layers 5 km 500 m
Cloud Fraction 5 km 500 m
Cloud top and base 5 km 100-500 m 50 m
Cloud Morphology 5 km 500 m photographs, etc.
-stratus/stratiform
-embedded convection
-layering/convection
RADIATION: Aircraft-based:
Parameter Horizontal Vertical Absolute Expected
Resolution Resolution Accuracy Values
Broadband Long-Wave Flux
(4-50 microns) 1 km 1-3 km 10 W/m2 100-400 W/m2
Narrowband Long-Wave Flux
(8-12 or 10-12microns)1 km 1-3 km 10 W/m2 100-300 W/m2
Broadband Short-Wave Flux
(0-5 microns) 1 km 1-3 km 2% 0/1100 W/m2
Narrowband Short-Wave Flux
(.55-.75 microns) 1 km 1-3 km 2% 0/1100 W/m2
Long-wave Radiance 1 km 1-3 km 2%
-narrow/broad-bands
Narrow-Band Radiance 1 km 1-3 km 2%
-visible/near IR
angular variation
RADIATION: Satellite-based
GOES (old)
VIS 1 km
11 microns 8 km
GOES (new)#
-VIS, 4 km
-6.5 (water vapour) 8 km
-3.9, 11, 12 4 km
METEOSAT
-VIS, 6, 11 5 km
NOAA polar orbiting 1 km
-VIS, near-IR,3.9,
11, 12
-TOVS
DMSP
-SSMI (microwave) <50 km
-SSMT (microwave) <50 km
-imaging channels 0.6 km
(uncalibrated)
LANDSAT 35 m
# The 5 channels on the new GOES satellites are:
visible (0.55-0.75), water vapour (6.5-7),
and the followng IR windows; 4 (3.8-4.0), 11 (10.2-11.2), 12 (11.5-12.5).
The polar limit of useful data from these satellites is about 70 degrees.
MICROPHYSICS AND PRECIPITATION:
Parameter Horizontal Vertical Absolute Expected
Resolution Resolution Accuracy Values
Precipitation Particles
-occurrence & rates 1 km .2-1 km 10% 0/50 mm/h
-phase 1 km .2-1 km liquid/solid/mixed
-habits and sizes 1 km .2-1 km 10% 0/5 cm
-riming 1 km .2-1 km light/heavy
Mixed Phase Particles
-occurrence/size 1 km 200 m 10% 0/5 cm
-mass 10% 0/0.1 g
-mass fraction 10% 0/1
Cloud Liquid/Ice Water
-cloud top and base 1 km 25 m 10-20% 0/.3 g/m3
-deep convection 2 km 500 m 10-20% 0/2 g/m3
-embedded convection .5 km 200 m 10-20% 0/1 g/m3
-frontal surfaces .5 km 200 m 10-20% 0/2 g/m3
-stratiform internal 5 km 200 m 10-20% 0/1 g/m3
Cloud Particles
-droplet size distn 100 m 10-200 m 10% 0/100 microns
-crystal size distn 100 m 10-200 m 10% 0/100 microns
Aerosols
CN, CCN, IN 10 km 1 km 10-25% 0/104/cm3
TEMPORAL SAMPLING: every 6 h
-mesoscale
-precipitation
-microphysics
-radiation
- Next section: GCSS Working Group 3: Annual Reports
- Back to GCSS Working Group 3 (Layered and Frontal) page
- GEWEX Cloud System Study (GCSS) page
- International GEWEX page
