Review of the ARM SGP CART Site Measurement Capability Relative to the Needs of ARM's Scientific Working Groups June 30, 1998 Randy Peppler, SGP Associate Site Scientist
INTRODUCTION Now that we are in our sixth year of operation at the ARM SGP CART Site, it is a good time to again review the site's measurement capability. This was done previously (1995), as documented in the Science Plan for the Atmospheric Radiation Measurement Program (ARM), published in February 1996. Section 6 of that report ("Instrument Development") contains information about then current and planned measurement capabilities, while section 6.4 describes "Concerns and Unmet Needs". At that time, concerns and unmet needs were listed for clear-sky measurements/retrievals, radiosondes, Raman lidar operation, RASS operation, surface meteorological instruments, microwave radiometers, cloud-sky measurements/retrievals, and optimum strategies with current instruments. IRF Group measurement needs from that time are expressed in section 3.3 (listed under radiation, radiative properties, and cloud variables). The present document should be viewed as an update to the 1996 effort. It was decided to poll key scientists within ARM's scientific working groups to assess their level of satisfaction relative to both the current routine SGP measurement suite and the site's IOP measurement capability, including heretofore-unmet measurement needs. To begin the process, an e-mail memorandum was sent on April 1 to the chief scientist and main DSIT representative of each of the ARM scientific working groups. The working groups polled were:
[It was hoped that IRF concerns would be captured within the responses received from its shortwave and longwave sub-components.] Input was solicited from each group in the following
two areas:
Relative to the first area, answers to the following were sought:
Relative to the second area, it was explained that due to tightening budgets and the inherent difficulty in scheduling aircraft, the ARM infrastructure is beginning to plan IOPs over a two-year horizon. Thus, input was sought on:
What follows are the responses received, as of June 30, 1998. They are divided into two sections, based on the two areas of input sought. In some cases, it appears that a polling of members of the scientific working group was conducted by the chief scientist and DSIT contact to come up with a consensus response; in other cases, it appears that the opinions expressed are mainly those of the chief scientist. It is hoped that ARM scientists and management can use this information to help guide the SGP CART site down the most scientifically and economically sound pathway possible. At some point it would be useful to prioritize the needs expressed here. ADEQUACY OF THE CURRENT, ROUTINE SGP CART SITE MEASUREMENT SUITE FOR EACH SCIENTIFIC GROUP'S NEEDS CLOUD WORKING GROUP (Response coordinated by Jay Mace) The quality of the operational CART data stream is very good and, for the most part, fulfills the needs of the Cloud Working Group quite well. We would like to address several issues, however, that concern us regarding the data stream and suggest ways that the data stream could be improved. Central Facility radiosondes: Due to budget shortcomings, the number of radiosondes launched from the central facility has been reduced from five per day during working days to three. Since our goal is to develop cloud properties from the full time series, we do not limit ourselves to weekdays. We have been forced, therefore, to find other sources of sounding-like information (we are currently using RUC model output augmented by the microwave radiometer water paths). Several of the algorithms we are implementing operationally require radiometric calculations in the thermal infrared region. We are, therefore, quite sensitive to the vertical location of water vapor features. The model output often fails to capture such features and when it does, they tend to be substantially smoothed in the vertical, thus, biasing the calculations and causing errors in the cloud property retrievals. Optimally, we would prefer at least six soundings per day evenly spread throughout the diurnal cycle. This would capture the gross evolution of water vapor features. The previous launch schedule (five per day) was adequate on weekdays. The current launch schedule is not adequate by any means and is certainly leading to errors in the cloud properties retrieved from the data stream. Another reason for a higher frequency of central facility radiosonde launches is motivated by the need to estimate the vertical distribution of clouds over the entire CART site during SCM IOP's. It may be possible to use the sonde profiles of relative humidity combined with MMCR data to form a statistical relationship suitable to predict the occurrence of clouds at the boundary facilities. Once this relationship is established, it can be applied to past SCM IOP's. This has the potential for increasing the return on the investment of high frequency SCM sonde launches. To best determine this relationship, a higher frequency of sonde launches at the central facilities is necessary. Raman lidar: An alternative to increasing the number of radiosondes is to deliver a calibrated and operational water vapor product from the Raman lidar. This would allow for very good observations of the full troposphere during the night and the lower troposphere during the day. The daytime upper troposphere would have to be sampled by radiosondes. With an operational Raman lidar product, the old radiosonde launch schedule (five per day) would likely be adequate for cloud property retrievals. The Raman lidar can also be used as an aerosol and cloud sensor and would provide depolarization information on hydrometeors. The depolarization information would allow us to distinguish the phase of the scatterers. Currently, we have no rigorous technique for distinguishing between supercooled liquid clouds and thin ice or mixed phased clouds. Cloud base observations: The Belfort Laser Ceilometer (BLC) is currently the only operational instrument that provides high vertical resolution cloud base observations, and a great deal of uncertainty still exists (five years after its installation) relative to what this instrument is reporting. At the Science Team Meeting in March 1998, a mean bias of 150m was reported between the micropulse lidar (MPL) and the BLC. We do not know if the BLC algorithm reports the height of the peak return, some unknown threshold value, or if it is just continuously in error. We are aware of an effort by Connor Flynn to bring this issue to resolution, but Connor is quite busy and it is unclear when this resolution will occur. We see three possible solutions: 1) understand and fix the cloud base algorithm of the BLC, 2) install a high resolution MPL at the central facility as soon as possible, or 3) install one of the Vaisala ceilometers as soon as possible. We would also like to see ceilometers installed at the boundary facilities as soon as possible. The problem of obtaining vertical profiles of cloud properties over the entire CART domain is a difficult one. Ceilometers will help substantially. MMCR: The MMCR data stream is contaminated much of the year in the lower troposphere by non hydrometeor targets such as insects and wind blown material raised from the surface. This contamination has been dubbed "atmospheric plankton". These non-cloud targets often bias the reflectivity or even mask the occurrence of weakly reflective boundary layer clouds. The MMCR is also limited by the processing power of the computers running it. The bug issue is being addressed but no solution is immediately at hand. The processing computer will likely be upgraded in the future, however, we are aware of no set schedule for this. Since Radian now has some control of the MMCR, we are quite concerned that commercial considerations will dominate over scientific concerns and that the upgrade will be delayed until Radian can be sure of profit for the effort. Furthermore, we would like to begin thinking of changing the fundamental data product of the MMCR from the moments of the Doppler spectrum to the Doppler spectrum itself. This would certainly increase the data flow from MMCR by a factor of many and would impact the Site Data System. We are now considering development of algorithms to extract significant return from the spectra. This is a long-term goal but we would like to move in this direction in the next year or two. Also, some uncertainty exists regarding the effect of the radome on the MMCR observations. The tarp, when it is wet, may cause errors. This effect needs to be quantified and fixed if necessary. Other issues regarding calibration are currently being addressed. Whole sky imager: This instrument has the potential of being a major component of our operational retrieval process. We are aware of significant progress made in making the data stream operational, but it has not happened yet. We would like to underscore the need for this data stream. Specifically, we need a continuous time series of cloud cover once per minute to match the nominal resolution of other cloud-sensing instruments. This is also an important element of the solution for determining how to best use zenith measurements to estimate area-averaged quantities and to validate the Minnis satellite estimates of cloud cover. The calibrated imagery will also be very useful and must be made a part of the WSI deliverable product suite. Microwave radiometer: The accuracy of the microwave radiometer is of the utmost importance in retrieving the properties of boundary layer clouds. Due to the statistical nature of the retrieval algorithm currently being used to convert the MWR radiances to liquid and vapor, a bias may exist in the products that severely limits our ability to determine cloud properties with certainty. We are aware of an improved algorithm constructed by Jim Liljegren that would substantially reduce the bias. We suggest that this algorithm be implemented on the operational data streams as soon as possible and that all retrospective data be reprocessed. It is currently being tested at the Morris (BF5) boundary facility. Near infrared radiometer: A need has existed for some time for a zenith viewing near infrared radiometer. This instrument will be useful for many purposes. However, an immediate need exists for solar radiation observations in the vertical column that the MMCR, AERI, MPL, and Raman lidar observe for retrieval validation purposes. [The NFOV infrared radiometer is scheduled for deployment at the central facility in August 1998.] Cloud observations away from the central facility: The SCM group requires some estimate of cloud fraction over the entire CART domain. We currently have no reliable means of generating this information from surface-based data. A possible solution to this is to install a hemispheric sky imager (HIS) at each extended and boundary facility. The HSI is a reasonably inexpensive alternative to the WSI and would provide the cloud fraction information required. Combined with the AERIs and ceilometers to be installed at the boundary sites, our ability to define the boundary conditions needed by the SCM group would certainly improve. Other elements of a program to estimate CART-wide cloudiness have already been mentioned. Chuck Long of NOAA/ARL/SRRB is developing the HSI, and has deployed a version of the instrument at the central facility during recent IOPs, including the fall 1997 Integrated IOP. Additional input from Eugene Clothiaux: It would be nice to a make sure that the MPL window does not get dew on it on nights when no low stratus are present. This occurred quite frequently early on and I believe I still see evidence for it. For example, on 16 January 1998, dew wiped out the detection of cirrus throughout the nighttime hours. Additional input from Martin Platt: We would like to see the high resolution MPL installed as soon as possible. This will enable quality cloud data up to the tropopause. The BLC only operates up to 7.3 km. A high-resolution MPL will provide better monitoring of cirrus, complementing the MMCR measurements. A dual-polarization, high-resolution MPL like the one that Jim Spinhirne has been designing would also allow phase discrimination between water and ice particles, as suggested for the Raman lidar, but with possibly increased reliability. The high-resolution MPL would provide:
The dual-polarization, high-resolution MPL would also provide:
A zenith-looking fast visible radiometer is also something that we would like to see at the site. Similarly, a spectral fast near-infrared radiometer would be beneficial. The Heimann radiometer should be checked out again against another sensitive, fast infrared filter radiometer. From previous tests, the Heimann does not detect below about -50° C brightness temperature, which is a disadvantage for the SGP and NSA sites, but is satisfactory for the TWP. We have a fast filter radiometer, constructed with funding from the ARM program. It was constructed for lidar/radiometer retrievals and was used successfully in Australia in the 1995 MCTEX. It is very sensitive and will detect clouds and water vapor under very cold, clear conditions (-80° C temperatures), as shown in the 1994 Spring IOP. A new Sterling cycle-cooled radiometer of similar design has now been completed, which can operate continuously without liquid nitrogen. It would be an advantage to install one at the central facility. The radiometer has a time constant of about 1-second, and would complement the AERI in picking up rapid radiance fluctuations in broken-cloud conditions. SHORTWAVE RADIATION WORKING GROUP (Response from Warren Wiscombe) We agree with the Cloud Working Group that the SGP CART datastream is serving us fairly well at this point. We just need to fill holes, get the shortwave spectrometers working as reliably as the AERI, and substantially improve our spectrometer calibration. General Points: Central Facility radiosondes: These are not nearly as important for the shortwave effort because even at five per day they would be far too infrequent to be of much use. Water vapor: Shortwave radiation at the surface is not that sensitive to the exact vertical profile of water vapor, as it responds mostly to the integrated water vapor column amount. Pressure: The only pressure value the shortwave group needs is surface pressure, to do Rayleigh scattering calculations. Temperature: The temperature profile has almost no effect on shortwave gaseous absorption calculations (the band strengths are only weakly temperature dependent within the range of temperatures where most water vapor is found), but it is important in showing places where the relative humidity climbs above 70 percent (where the fattening of aerosol particles due to hygroscopicity becomes important). We also need the temperature profile inside clouds, which coupled with the location of cloud base allows us to calculate the water vapor amount inside clouds (assuming saturation). However, we might be able to get by with just the surface temperature and assume the cloud base is at the LCL. Raman lidar: This is mainly of use to us in describing aerosols. While the aerosol backscatter is of some interest, we cannot use it nearly as easily (to input models) as the new aerosol extinction profile product offered by Rich Ferrare for June 1998. Extinction can be input directly into models. We also need the Raman lidar to identify regions of relative humidity above 70 percent, for the reasons given above, which are again aerosol-related. The actual water vapor profile only weakly affects shortwave calculations, within realistic limits. Cloud base observations: This is not as important in the shortwave as in the infrared. We can move the cloud base up and down a few hundred meters with almost no effect on radiation, all else remaining equal. We are much more sensitive to cloud optical depth. MMCR: We are as affected by the current MMCR bug/plankton and computer processor problems as the Cloud group. We desperately need to know the vertical profile of cloud liquid water and/or ice, and the MMCR is the ONLY source of such data for us. Admittedly, there are a number of assumptions in the MMCR processing to retrieve liquid water or ice, but this is a lot better than nothing. Thus, we suggest offering the atmospheric plankton problem to the Aerosol Working Group. This plankton fits the technical definition of aerosol and it does float without falling, so it cannot be called "hydrometeors" or even "meteors". It merely adds another mode to the aerosol size distribution, peaking around 1 mm. If Gulfstream-1 collection and analysis capabilities could be directed toward this problem, and perhaps the surface AOS as well, we think progress in resolving or at least understanding it would be more rapid. Whole Sky Imager: Our interest is very different than that of the Cloud group, but we would very much like to see a calibrated radiance as a function of angle coming from the WSI. The angular variation in shortwave models has barely been tested, and this would provide a start in that needed direction. Near infrared radiometer (NFOV): We also have a considerable need for this capability. It will allow us to more directly correlate radiation coming from the same field-of-view seen by the MWR and MMCR, sans the complicating influence of Rayleigh scattering. In a way, this also furthers our angular modeling goal relative to the WSI. Radiometry: Broadband radiometry: We are ecstatic about the RCF, which covers broadband nicely. We are also happy about the aforementioned NFOV radiometer with a field of view matching the MWR. We are not happy with the slow pace of GRAMS implementation. Some way needs to be found to accelerate GRAMS progress toward calibrated fluxes (it now reads only millivolts). Spectral radiometry: Except in the ultraviolet, there are no glaring unmet needs at the surface as long as we have a robust SWS and hopefully the SSP (Graeme Stephens). These are desperately needed as a crosscheck on the RSS, which is advertised as little more than a Modtran machine by its maker. The ASTI is not crucial at the present time for solving the enhanced absorption problem, but if the lower-resolution spectrometers turn up any clear-sky problems beyond 1-micron wavelength, then the ASTI can be brought to bear to try to pinpoint the problems. There are conflicting opinions about the ASTI. The IRF recommendations usually include a long recommendation about how we need the ASTI. Lately, the recommendation has been to extend the ASTI down to 0.7 microns from its present cutoff at 1 micron. This is portrayed as absolutely crucial to the success of ARM's shortwave efforts. This is unfortunately an exaggeration. The present spectrometers also cover the 0.7 to 1-micron region, with coarser resolution, and if they turn up significant discrepancies compared to models there, then certainly we need to extend the ASTI down into that region. But to do so without cause seems premature, considering the cost. So I would not rate further development of the ASTI crucial, especially considering its tenuous performance in recent field campaigns (in spite of major ARM funds applied toward its development). UV spectrometry: While claims are made that the RSS goes down to 0.36 microns, a more honest assessment is that all of the spectrometers are accurate only down to 0.38 to 0.40 microns. The signal-to-noise ratio of silicon detectors, which they all use, falls off rapidly below 0.40 microns. That leaves us somewhat blind, spectrally, between 0.3 and 0.4 microns (the broadband radiometers include this region). At the fall 1997 IOP, Tom Stoffel ran some UV radiometers, but they wouldn't be satisfactory for permanent installation because they do not in fact capture all the radiation from 0.3 to 0.4 microns (or 0.3 to 0.38 at least). They say they do, but when you read the fine print, they only capture narrow pass-bands and then extrapolate to get a broadband result. So what we need is a broadband UV radiometer than honestly measures from 0.3 to 0.4 microns wavelength. We would encourage the nascent effort to return calibrated radiances as a function of angle from the WSI. The poster on this at the ARM meeting was fascinating and showed distinct disagreements with modeling. The angular variation is one we have not explored, and provides another independent way to test models. Finally, we would like to see more angular information gathered from the Cimel sunphotometer (CSPHOT). It normally only scans 40 degrees on each side of the sun, but it has another operating mode that can do a complete azimuthal and zenith (principal plane) scan. We would like to see the CSPHOT operated in this alternate mode some of the time, perhaps once a week. Aerosol Optical Depth: The MFRSR and CSPHOT seem to be performing well in retrieving aerosol optical depth. Comparisons at the fall 1997 IOP indicated accuracies of 0.01, or about the limit of measurement capability at this time. Column Vapor and Liquid: We are highly dependent on the MWR for column vapor and liquid. Radiometric data, even spectra, are taken once a minute or faster, and we need these quantities on the same fast timescale. The improvements that Jim Liljegren is making to the MWR algorithms, as part of his Science Team effort, are sorely needed, and we look forward to their implementation. These improvements especially affect the liquid column, which is afflicted by spurious low values (some negative) in clear skies, which make the current datastream hard to work with. Based on interactions with Ed Westwater, we are concerned that some thresholds put into the MWR algorithm to produce good water vapor results have caused a non-physical cutoff in the liquid values. Our radiation measurements would be sensitive to large liquid values, and if the MWR is not retrieving them correctly, this could be misdiagnosed as model failure. This may not be a problem, but it does concern us. Finally, we would very much like to do some validation of the MWR liquid values in low stratiform clouds that can be overtopped by the tethered balloon (i.e., cloud tops below 1 km). No such validation has ever been done, surprisingly, because in situ column liquid water path is a devilishly difficult measurement. Aircraft obviously cannot do it. Our idea is to attach small hot-wire liquid water probes at intervals along the tether and actually measure the liquid water profile, whose vertical integral can be compared to the MWR retrieval. This is not as futuristic as it sounds - already AIR, INC., sells a tethersonde package with six instruments attached at intervals along the tether line. Radar: As discussed earlier, Shortwave Group cloud research is crippled by the bug/plankton problem. There are apparently ways to get rid of the bug returns below cloud base (this should be done as soon as possible), but the bug and droplet returns are indistinguishable INSIDE the clouds. Because of this, we can't trust the radar inferences of cloud structure for clouds below about 4 km, except in the winter. This is not fatal for the Shortwave IOP effort because we at least still have the MWR liquid water path, but it leaves us with a big loophole when trying to compare models and measurements (namely, how do we distribute the MWR liquid water in the vertical?). Presently, it is difficult to come up with solutions when we are not sure what the plankton is - what fraction is bugs, what fraction is spider webs and spiders, what fraction is seeds, what fraction is spores, and so on. LONGWAVE RADIATION WORKING GROUP No input was received. IRF WORKING GROUP Recommendations from the 1997 IRF Meeting that may or may not still be of relevance are listed:
Additional action items from notes compiled at the 1998 IRF Meeting (not already covered by the 1997 recommendations):
AEROSOL WORKING GROUP (response coordinated by Meng-Dawn Cheng) With regard to the aerosol data collected at the SGP CART site, Steve Schwartz of BNL would like to see collection of aerosol chemical composition and aerosol relative humidity data. Within ARM now, discussions are ongoing concerning 1) the deployment of a nephelometer on the 60-m tower to get some sense of the vertical profile of aerosols from the AOS to 60-m, and 2) the possibility of routine (several times per week) in situ measurements of boundary layer aerosols over the central facility using a small aircraft. Joyce Penner of the University of Michigan indicates the need for CCN spectrum and for CCN spectrum plus aerosol composition as a function of size during IOPs. Her group is working on algorithm development and will not know if their model and assumptions are correct until all of the satellite data since April 1996 are available. Tad Anderson of Robert Charlson's University of Washington group has developed a new 180-degree backscatter nephelometer and is ready to begin testing it. The instrument basically consists of a standard 1-wavelength TSI nephelometer with a laser added to get the 180-degree scattering coefficient. It may be possible to operate the instrument to measure total scatter/hemispheric backscatter and alternately lidar-backscatter. The laser wavelength is around 520-530 nm. Plans are to test it in Seattle and on the Washington coast. Anderson is looking for an aircraft platform to test it on. The next Aerosol IOP might be a possibility. Tom Charlock provided the following material:
CERES cloud recommendations: There is a need for continuous lidar/radar/radiometer vertical profiles of cloud microphysics at all three ARM sites with a time sampling roughly 1 minute. These data would be used to validate CERES cloud retrievals using thousands of surface/satellite matches. There is a need for many additional in-situ aircraft microphysical observations over a complete range of water and ice cloud types at the SGP site. These would be used to validate the ARM cloud profile retrievals. A "complete range of water and ice cloud types" means all cloud conditions to which the radar/lidar/radiometer cloud profile retrievals are sensitive; i.e.:
For cloud fractional coverage, scanning lidar is the optimal system. Nadir pointing lidar is extremely useful but more difficult to match to satellite areal data. A stopgap may be to combine all-sky cloud fraction (has the problem of missing thin clouds) with the direct/diffuse estimate of cloud occurrence (sees very thin extensive clouds but cannot distinguish areal coverage of thin versus thick). All-sky plus direct/diffuse may provide the best areal average cloud fraction strategy. It is suggested that discussions take place with Chuck Long at NOAA on the status of these two methods and possibility to combine them. As in the first item above, for cloud microphysical properties, continuous data is needed. After validation of cloud vertical profiles using lidar/radar/radiometers, we will need a scanning version of nadir cloud retrievals to do three-dimensional cloud volumes. This will allow us to scan perpendicular to the wind direction (a la Ed Eloranta's scanning lidar data) and produce cloud volume measurements. These will be optimal for studying aircraft radiative flux measurements as well as surface fluxes to remove problems caused by three-dimensional cloud spatial inhomogeneity. This should be a long-range goal. CERES aerosol and flux recommendations: Continue the present SGP measurements of aerosol optical depth (MFRSR and CSPHOT), broadband radiative flux (SIRS and GRAMS), height profile of aerosol scattering (MPL), and height profile of water vapor and aerosols (Raman lidar). We have not used CSPHOT or Raman lidar data yet, but we will. Jim Spinhirne gave us estimates of aerosol height profiles based on the MPL, but these quite useful height profiles may not be standard ARM products. CERES is very eager to use the Raman lidar data. As always, CERES has a great interest in accurate measurements, especially for the broadband fluxes. Develop continuous aerosol optical depth measurements (a) covering the near infrared and ultraviolet, as well as the visible, and (b) sampling the shortwave more densely in wavelength than present measurements (mostly of the visible) by MFRSR and CSPHOT. This calls for a direct/diffuse spectrometer or a radiance/scanning spectrometer. It would be valuable for aerosol and cloud investigations. Measure aerosol single scattering albedo (SSA) at more wavelengths (beyond the present SSA at 0.55 micron from the surface-based ASOS), continuous at surface with vertical profiles from aircraft during field campaigns. Deploy a new system for the measurement of aerosol SSA. There are doubts about present methods, which involve filter impaction. Are spectrophone measurements of aerosol absorption possible? Enhance processing by ARM of aerosol and cloud related measurements. For example, inversions for aerosol size distribution from MFRSR and CSPHOT data, cloud optical depth retrievals from MFRSR, and cloud screening from combinations of lidar, WSI, and radiometer data would be desirable. CERES surface optics and the use of field campaigns: The current level of observation of in situ measurements of aerosol size distribution is appropriate and should be maintained. We plan to use such measurements, both at the surface and from aircraft, during the August 1998 summer NASA helicopter BRDF campaign. It has been suggested to us that the current in situ measurements of aerosol size distribution at SGP are not that accurate, and that for aerosol size distribution, we should instead use inversions from the MFRSR or CSPHOT. But, we are aware that those inversions do not use input from the near infrared, so the question arises whether those inversions are thorough and accurate. Also, in situ measurements of aerosol size distribution may be the only way to observe special features such as a thin layer of large dust particles that might be very near the surface. WATER VAPOR (Atmospheric State) WORKING GROUP (Dave Turner, Hank Revercomb, Dave Tobin, Bob Knuteson) The completion of two of the three planned Water Vapor IOPs has greatly improved our understanding of the characteristics of the current SGP water vapor measurement capabilities. There are six primary water vapor measurement systems in routine operation at the SGP CART site - MWR, Raman lidar, in-situ measurements (60-m tower, SMOS, THWAPS), AERI, BBSS, and GPS. These are addressed in turn below, with respect to a) their role as an absolute standard for water vapor measurements and b) profiling capability. Microwave Radiometer (MWR): This instrument has the potential to be an absolute standard, but some questions about microwave calibration have been raised. We endorse the current activities undertaken by Jim Liljegren to allow TIP curves to be taken routinely and used to automatically monitor (and update) calibration. Periodic intercomparisons with other SGP MWRs should continue, as well as comparisons with guest radiometers on an IOP basis. Raman lidar: This instrument has shown an ability to make the needed high resolution (spatially and temporally) water vapor measurements at the central facility that are needed by the IRF group. The two remaining issues with this instrument are its long downtime periods and continuing nonlinear calibration issues (namely the pulse-pileup and overlap corrections). Both of these need to have high priority from the lidar instrument mentor, DSIT representative, and members of the Science Team. The best way to provide an absolute calibration for these profiles remains an area of active research and is discussed below. Comparisons with the NASA GSFC scanning Raman lidar on an IOP basis are invaluable. In-situ measurements (60-m tower, SMOS, and THWAPS): The new Vaisala humidity sensors mounted on the 60-m tower at 25 and 60 meters prior to the fall 1997 Water Vapor IOP demonstrated remarkable agreement with co-located chilled mirror hygrometers. This result leads us to believe that the Vaisala probes can be used as a calibration standard, but care must be taken to maintain their calibration. Having redundant sensors at both levels (and at the THWAPS) is a plus, as is the planned temperature and humidity calibration facility at the SGP site. Comparisons with well-calibrated independent sensors should continue on an IOP basis. AERI: The AERI radiance spectra are a key aspect of the Water Vapor IOPs with respect to their role in the AERI/LBLRTM QME. This QME quantifies how well we can calculate downwelling infrared radiance and is used to monitor and evaluate the stability and calibration of various water vapor and temperature measurements. AERI + GOES water vapor/temperature retrievals: This technique provides our best avenue to continuously profile water vapor and temperature during clear sky conditions at the boundary facilities, but improvement (using a refined forward model for the AERI retrievals, along with polar orbiting satellite instruments and GOES) and validation of the technique still need to occur. Some of this validation can be performed with historical data at the central facility during Water Vapor IOPs. The AERI + GOES retrievals will operate in partly cloudy conditions as well. Radiosondes: The recently reduced number of soundings during non-IOP periods does impact our focus area, as sondes will continue to be the workhorse in and above clouds and in the upper troposphere during the day, as well as serving as a common transfer standard to compare the different water vapor measurements. As some of the calibration issues of the Raman lidar are worked out and advancements are seen in the AERI+GOES water vapor/temperature retrieval, the need for radiosondes will indeed be reduced. However, the three launches currently being made (0000, 1200, and 2030 UTC during the week) hamper our ability to fully validate both the Raman lidar and the AERI+GOES retrievals due to the uneven distribution of samples. We would move to launch four sondes per day at equal intervals (say 0000, 0600, 1200, 1800 UTC). The 0600 UTC launch would provide a nighttime sounding. Eight launches per day during SCM IOPs are still desired, and will help improve statistics. Barry Lesht has pursued the possibility of selecting radiosonde batches from Vaisala for the SGP CART site that are determined to be of the highest quality. We endorse this activity and recommend that the necessary steps be taken to see that this happens. GPS: The Lamont GPS site is providing routine precipitable water vapor measurements. During the 1997 Water Vapor IOP, we saw that its precipitable water vapor values were nearly identical to values derived from a NOAA GPS unit deployed at the SGP site (the two units having a physical separation of about 9-km). Thus, we do not see the need for a second unit to be installed permanently at the SGP site. During IOPs, however, it would beneficial to have a second, redundant unit at the SGP site. The GPS has the potential to be considered as an absolute water vapor standard, although improvements in the precipitable water vapor processing algorithms are required. The critical ARM instruments to the success of the Water Vapor Working Group are the MWR, 60-m tower measurements, the Raman lidar, and the AERI. The sondes will continue to play a very important role as the calibration issues of the lidar are worked out, and afterwards for profiling during cloudy sky conditions. AERI and sondes are also crucial for temperature profiles, while high altitude sonde (and perhaps AERI + GOES observation) temperatures will be essential for combining upper tropospheric moisture measurements with radiance observations. The WV group is continuing to study ways to use the "absolute standards" (the MWR, 60-m tower measurements, and GPS) to help reduce the bias and variability in the profiles measured by both the Raman lidar and the radiosondes. One of the primary questions that still needs to be addressed is whether scaling the profiles to agree with the tower at a point (60-m) or with precipitable water vapor from the MWR is the correct action, as both techniques weight the lowest portion of the profile heavily. SURFACE FLUX WORKING GROUP (Response from Chris Doran, with an addition from Randy Peppler) The big issue with the surface flux group is the establishment of a reliable set of ECOR measurements for the both sensible and latent heat fluxes. There have been requests for supporting data, such as hand-held NDVI readings at the EBBR and ECOR sites - the status of this possibility is unknown. Regarding ECOR, it may well be that establishing a half dozen reliable sites would be better than a dozen that operate only occasionally or are difficult to keep running. The establishment of an IDPC for the extended facility ECORs has been given a higher priority within the SDS group. Work should begin this summer. If we can begin to ingest data from the remote ECORs, we may then be able to get a better handle on the ECOR sensor malfunction problem. Continuous ECOR measurements are necessary to obtain data through a full crop growing and harvesting cycle, along with data from bare fields when no crops are present. SINGLE-COLUMN MODEL WORKING GROUP (Response coordinated by Ric Cederwall) This information is extracted from our discussions at the most recent SCM Workshop. The SCM researchers require data sets that are often at the highest level of value adding, and quite removed from instrument-level data streams (we are at the end of the data food chain). For this reason, our data needs may not translate directly to a given measurement, but rather to a series of value-added data products. The other working groups will undoubtedly deal with more instrument-specific data streams. The SCM benefits from high quality measurements, so we expect we would endorse many of the recommendations put forth by the other groups. Clouds: Basic cloud characterization across the SGP site is needed to estimate cloud fraction as a function of height and time. The MMCR, and associated cloud products, provide great detail at the central facility. We look to the Cloud Working Group to provide their best estimate of macrophysical cloud characterization across the SGP site. To accomplish the above, additional cloud detecting capabilities are needed at locations other than the central facility. Ceilometers at the boundary facilities will help, along with some form of cloud imaging there (i.e. the HSI). Satellite-based products help extend surface-based point or limited area data to form site-wide estimates of cloud characterization. Besides cloud fraction, the additional cloud properties critical for the SCM include cloud boundaries, cloud overlap, particle size, cloud droplet and cloud ice number concentration, cloud optical thickness, and horizontal advective tendencies of liquid water and ice. Atmospheric state (see also discussion under SCM IOPs below): The SCM Working Group seeks to use atmospheric state profiles obtained from sources other than radiosondes. Data to support SCM research have been confined to three-week periods (SCM IOPs) due to the high costs of the three-hourly soundings at the five launch locations. SCM research will benefit from having longer periods of data needed to force the SCMs and evaluate parameterization performance. Many of the evaluation approaches are statistical in nature, since the GCM parameterizations themselves are based on ensemble characterization of the phenomena addressed. Although short-term case studies have assisted the SCM Working Group in their initial work, we have too few of such case studies to support substantive conclusions about parameterizations tested. Development of retrieval algorithms for temperature, water vapor, and winds from remote sensing has been an ongoing effort in ARM. We support what measurements are needed to obtain the best possible retrievals of the atmospheric state in the column over the SGP site that we now obtain from radiosondes. The Atmospheric State Steering Committee in the IRF Working Group is tasked with guiding this effort, and is the best source of measurement requirements for the SGP CART site. Surface flux over cropland: We need more consistent sampling of surface fluxes over cropland. The current surface flux dataset is predominantly from the EBBR stations, which sample undisturbed areas such as pastures and rangeland. The ECOR stations sample the croplands. The ECOR systems have produced much less data due to instrument system reliability problems, which are inherent in this research level system. Nonetheless, the data sets expected from the ECOR systems are of high priority for SCM surface forcing, where they are used in (i) direct fusion of the point observations into site-wide values of surface flux, and (ii) validating the SiB2 model-based flux estimates provided by Chris Doran. It is clear from the averaged summer values for EBBR that we are missing the hotter, drier surface conditions on clear days over the harvested wheat fields concentrated in the central north-south band of the SGP CART site. IOP IDEAS/PLANS FOR 1999 and 2000 CLOUD WORKING GROUP (Response coordinated by Jay Mace) The Cloud Working Group would like to hold another Cloud IOP in the April/May 2000 timeframe. This will give us sufficient time to analyze data already collected and to develop a reasonable scientific plan. This IOP should include both the Wyoming King Air and the North Dakota Citation since we will want to examine multi-layer and deep cloud systems. There is also a need to image ice crystals within a wide size range from ten to several thousand microns. Since spring 2000 will put us well into the EOS era, some coordination with NASA should be planned. This may allow for a more thorough experiment that could include participation by the ER-2 and the DC-8. SHORTWAVE RADIATION WORKING GROUP (Response from Warren Wiscombe) Before the opportunity arose to coordinate a second Shortwave IOP with the NASA-Langley BDRF campaign in August 1998, a desire was indicated to hold another IOP in spring 1999 or 2000, perhaps in March or early April. The springtime period would allow for a better chance of studying stratiform clouds. Winter weather likely would be too rough and guest instruments might not be rugged enough to stand the beating. March is probably the earliest that the weather is not still awful and yet provides lots of cloud opportunities. The August 1998 IOP, like the first Shortwave IOP, will primarily focus on clear skies. More detailed future plans will be made after this summer's IOP. LONGWAVE RADIATION WORKING GROUP (From Doug Sisterson and Randy Peppler) It has been learned that Joe Michalsky and Tom Stoffel are organizing an international pyrgeometer comparison for the September 1999 period. It would be held at the central facility. AEROSOL WORKING GROUP (Response coordinated by Meng-Dawn Cheng) Peter Daum and Steve Schwartz of BNL are interested in a wintertime study of cloud and aerosol properties. All previous Cloud/Aerosol IOPs have been in either fall or spring (and soon summer for aerosol) when clouds and aerosol vertical distributions are fairly complex. Winter clouds tend to be much more stratus-like and are therefore much easier to characterize and work with. Thus, they suggest that ARM consider a winter IOP, possibly in 1999 or 2000. With regard to IOP plans in 1999 or 2000, the Penner group similarly thinks that the season with the most widespread low level cloudiness is the one when aircraft with aerosol measurements and a full suite of cloud data instruments should be taken. WATER VAPOR (Atmospheric State) WORKING GROUP (Dave Turner, Hank Revercomb, Dave Tobin, Bob Knuteson) Future Water Vapor IOPs will concentrate on upper tropospheric measurements of water vapor, as well as vertical distributions in the lowest portion of the atmosphere (both with aircraft and with chilled mirror hygrometers on a tethersonde/kite). To that end, dewpoint and frost-point hygrometer data from aircraft during future IOPs is highly desirable. This focus on upper level moisture is an important issue with respect to using the SGP CART site as a satellite validation site. For the next IOP (1999), we suggest an IOP that would focus on 1) the absolute calibration issue, 2) upper tropospheric measurements, and 3) an international radiosonde intercomparison. The major constraints on scheduling this IOP are a) combining it with other IOPs (i.e., SCM IOP) to get the required sonde launches, b) coordination with DIAL systems and aircraft based radiance observations, and c) the participation of other radiosonde sources. One specific plan is to bring a CO2 DIAL to the central facility to run side-by-side with the SGP Raman lidar (this was once scheduled for the fall of 1998, but has been delayed indefinitely). This opportunity would offer a unique chance to compare diurnal water vapor measurements made by both lidar systems and to help study the possible diurnal characteristics of the sondes that have been noted previously. For the upper tropospheric focus, we would like to have in situ measurements as well as LASE (NASA) and, if possible, HIS-like measurements (HIS = High resolution Interferometer Sounder). The HIS is designed to be flown in an airplane to provide high resolution longwave radiance data from high altitude. It was the Wisconsin forerunner of the AERI. At present, we hope to fly the LASE and a scanning HIS (the next generation of the HIS) on the NASA DC-8 and to combine this with a (currently unspecified) previously planned experiment to minimize cost. We should also continue to compare the CART instruments in detail with outside measurements, and hence encourage the participation of as many invited groups as possible. A non-exhaustive list of possible guest instruments is given below. For the year 2000, we are looking into the possibility of conducting an Arctic Water Vapor IOP, which would focus on the measurement of water vapor under cold, dry conditions. For this IOP, we would also like a focus on incorporating various satellite-based observations. Possible IOP guest instruments and procedures and some of their roles would include:
SURFACE FLUX WORKING GROUP There are no plans at present. SINGLE-COLUMN MODEL WORKING GROUP (Response coordinated by Ric Cederwall) For fiscal year 1999, the SCM Working Group recommends that there be three SCM IOPs:
For fiscal year 2000, the working group has discussed the option of driving SCMs with retrievals of atmospheric state from remote sensing. We expect that by FY2000, ARM will have made substantial progress on such retrievals. We further expect that some amount of radiosonde launches will be needed by the retrievals to achieve desired accuracy, and that the success of the retrievals may be seasonally dependent, based on representativeness issues. If this is the case, then we recommend that the sonde-launching schedule be consistent with the retrievals and spread across longer periods than three-week IOPs to support more SCM run days. If the retrievals are not adequate in FY2000 for supporting SCMs, then we recommend that there be three SCM IOPs, with timing to be specified in the future by the SCM Working Group. Scheduling of SCM IOPs with other IOPs is highly desirable, especially those collecting cloud and radiation data in the column. UAV GROUP No input was received. |