Wayman Baker (NOAA/NCEP) called the meeting to order, welcomed the participants to North Conway, and led a review the Action Items (AIs) from the Previous (January 2002) meeting in Key West, FL. The status of each previous action items is summarized below:
Wayman Baker reported on the Status of the NASA New Millennium Program for Chris Stevens. The immediate challenge of NMP is to deal with budgetary shortfalls in NASA. Probably the latter part of FY05 will be reached before new starts in NMP are entertained. On a positive note, NMP is partnering with the Space Test Program (STP) for other missions and it is in purview of NMP to do so for a DWL mission as well.
Pierre Flamant (CNES) presented an update on ESAís Aeolus-Atmospheric Dynamics Mission. ADM entered Phase B development on July 1 2002. Laser (transmitter) development continues in Phase B with two competing teams. Pierre expressed gratitude for the input of the ADM advisory group, but noted that they have had little information regarding the transmitter since extra care must be taken to avoid the possibility of leaks while still in a competitive phase. He noted that one team started 6 months before the other. Pierre reviewed the ALADIN pre-development model activity, and the instrument and mission implementation concept, and data specification - 1-2 m/s for any LOSH (depending on height) with no bias and no correlation errors. Pierre acknowledged that the 150 mJ tripled YAG at 100 Hz PRF is a very challenging specification. The contractor, THALE, has a diode-pumped pulsed high-energy air-cooled UV laser for which he showed a block diagram including Q-switched oscillator and cooling via three heat pipes. A plot of oscillator energy versus optical energy indicated 7% efficiency. Power amplifiers are still being worked on and no results were shown. The pump laser diode assessment: shows that 2000-hour endurance has been achieved. All critical units have been designed and released for manufacturing. ADM science studies continue to progress - two on signal processing, for stationary and unstationary conditions, respectively, and a study on error correlation by KNMI. The signal processing studies considered various interferometer (Fizeau, Mach-Zehnder) possibilities to process Mie returns. A virtual instrument simulator has been developed using neural networking. VALID 2 and 3 field campaigns building on the initial VALID (1999) exercise have been planned in support of ADM. An AEOLUS Ground campaign is planned for four weeks at Lindenberg in Germany, as well as an AEOLUS airborne campaign intended to verify the Level 1b algorithm and proposed QC schemes. Support is expected from ESTEC. Paul Hays asked about the estimated cost at launch for ADM. Pierre replied $250 M Euro. Dave Emmitt noted that the 25% duty cycle implies switching on/off, to which instrument specialists always express aversion. Pierre replied that the seeder is ALWAYS on, only the master oscillator (MO) is cycled on/off, and the MO looks okay so far in tests. Bruce Gentry wanted to know whether the 25% is to be constant, or whether targeted observations as well would be allowed as long the average did not exceed 25%? Pierre confirmed that regular, not targeted observations are specified. Asked when the Lindenberg test will occur, he indicated perhaps in 2004. In response to other questions, Pierre confirmed that a 3-year on-orbit lifetime is specified.
Kohei Mizutani (Communications Research Laboratory) reported on the Space-borne Coherent Doppler Lidar Program in Japan. He reviewed the planned accommodations for the JEM on the ISS, and how this instrument concept is to provide tropospheric wind data from space during a 1-year non-operational mission. He reiterated that the JEM CDWL program faces schedule uncertainty now owing to the on-going merger of three agencies in Japan.
Michiko Masutani (NOAA/NCEP) provided an update on the NCEP global OSSEs, noting that this first calibrated OSSE using the NCEP data assimilation system (DAS) and based on the ECMWF Nature Run (NR) has been a joint effort by NCEP, NESDIS, the NASA DAO, Simpson Weather Associates, and an external Advisory Group. The goal is to evaluate minimal and optimal DWL product impact in conjunction with other new instruments expected in the same time frame. Recent progress includes upgrading AIRS data simulation for the NCEP Spectral statistical Interpolation (SSI) objective analysis scheme, evaluation of surface data, testing various errors for surface data, maintenance of the current SSI, and coordinated simulation of Cloud Motion Vectors (CMVs) with NESDIS CIMMS. Improvements include a new radiative transfer model, improved bias correction for radiance data, upgraded background error covariance, the assimilation of precipitation, and consideration of new instruments - notably AIRS and DWL. Expected benefits extend beyond instrument evaluation to preparing for real data and incorporating them into operations as quickly as possible. Michiko described the ECMWF NR, and revisited the shortcoming of using constant sea surface temperatures (SST) - this has been shown to be a significant departure from using realistically variable SST, as determined by TOVS, causing the impact of all satellite data, including a DWL, to have a smaller than expected impact. She proceeded to review the four bracketing OSSE experiments run to determine whether there is enough difference between extremes to justify closer examination of the intermediate sampling scenarios through subsequent OSSEs. Global anomaly correlation statistics show considerable impact from a DWL. Dave Emmitt commented that some of the recent results show that boundary layer measurements are important, in particular, where there are clouds. Future plans are to evaluate the DWL impact with new surface data and to add cloud-motion wind data and AIRS. Possible further work includes refined data sampling and density for DWL (i.e., other than the bracket values), reviewing error assignment, and exploring DWL adaptive targeting strategies.
Mike Hardesty (NOAA/ETL) presented, for FSL, the corresponding update on the regional-scale OSSEs, for which NOAA/FSL has done a regional NR, while NCAR is assessing model results including the NR, and the Global Background is provided by NCEP and SWA. The Regional OSSEs are expected to benefit from DWL data that have been assimilated into a global model, because improved global winds in turn improve the boundary conditions (BC) for the regional model. Since the regional model has a shorter assimilation time, additional impact might come from using DWL within the regional model; thus, there are two means by which DWL might be expected to benefit the regional model. Mike reviewed the methodology, including extraction of a suite of observations from the NR, and discussed the model physics and domain for the regional MM5 NR. A control experiment with no BC improvements (from the global model) and no DWL in the regional domain was run to provide a basis for comparison with improved BCs alone (to assess how much impact came from global model assimilation), and with both improved BCs and DWL data. FSL has assumed ìidealî DWL for the first regional OSSE. A realistic model of lidar and clouds is planned for later. The instrument model is the idealized direct detection fringe imaging instrument in low orbit. Will also incorporate a cloud model in the future. A question was raised about the benefit of doing these OSSEs for a data rich region. Mike acknowledged that this is a concern, and noted that with the tools developed, one could apply the technique to another (data poor) region. Dave Bowdle asked what metrics will be used for the regional OSSEs. Mike indicated forecast skill scores are to be the first-cut metric, and noted that having gone to small scales, FSL wants to progress to metrics that measure how well cyclone features are resolved. A final comment was that the particular NR did not contain especially interesting phenomena, and recommended that subsequent regional OSSEs be conducted with another NR. One that might investigate compelling cases such as a northeastern (USA) snowstorm.
Mark Phillips (Coherent Technologies Incorporated) reported on the ACLAIM/CAMEX 2-micron aerosol backscatter experiment. He noted that the previously existing 2-micron backscatter database lacks high altitude data confidence, which this effort was intended to rectify. He presented overviews for a variety of flights taken at the end of 2001. The ACLAIM system specs (8 mJ at 12.5% efficiency with a 10-cm telescope) were summarized. Turbulence effects were considered to be negligible at flight level, as was atmospheric extinction. The result: of the 40 h analyzed, 5 h were stored in a database, consistent with ground-based measurements. The results are more conservative than the NASA/GLOBE aerosol model; some model results that have appeared in Applied Optics seem in accord with the ACLAIM results. Dave Emmitt noted that it is encouraging that the reference atmosphere based on lidar, optical counters, and models all fall consistently to the background aerosol level. Dave Bowdle asked how the location of the sensor might have degraded the measurements as Mark had alluded. Mark replied that the early analysis seems to indicate that it is desirable to have the sensor close to the front of the aircraft.
Bob Atlas (NASA/GSFC Data Assimilation Office, DAO) presented initial results from OSSEs based on a new NR from a finite volume (FVGCM). The USWRP seeks to mitigate the effects of extreme events (hurricane landfall; air pollution episodes, etc.) and NASA is responding with new observational systems as well as improved NWP models and data assimilation methodology, especially for satellite data. Major improvements are sought out in forecast accuracy by 2010 and again by 2025. Though it is uncertain whether the government will support this goal, Bob argued that it cannot be reached without DWL observations, and more OSSEs to determine what would be a good experiment or instrument. Previous (1985) DAO OSSEs indicated the potential for DWL. But the advancement in data assimilation is so rapid, one canít assume the DWL will still have impact, especially with more competing data. This provides motivation for this work and the on-going NCEP OSSEs. Bob noted that the ECMWF T213 NR has a couple of deficiencies - the lack of realistic fronts and cyclone evolution, and no hurricanes, which the DAO has sought to address by generating its own NR based on the FVGCM model, which conserves mass, momentum, energy, via a terrain-following Lagrangian (grid point) scheme. It is more accurate for transport, and as computationally efficient now as spectral models, which formerly had an advantage. The FVGCM appears to do an outstanding job representing fronts and cyclone, and can simulate a hurricane with an eye, eye wall, spiral bands of precipitation, and reasonable 3-D (vertical) cross section, too, even though _ degree resolution leads to weaker hurricanes than in reality. DAO has also generated a 100-day NR from the GEOS model. The new NR passes the test for realism, including cyclone verification and cloud patterns, although adjustments are needed. For the OSSEs, a different model is used for the data assimilation than for the forecast-GEOS 3. The OSSE system has been successfully calibrated by comparing the impact of simulated observations to that achieved by using the corresponding real observations. Forecast experiments with ìrealî data include: Control, Control plus satellite temperature, control plus QUIKSCAT. Simulated data experiments include DWL and AIRS. DWL data simulated at TOVS locations with 1 m/s accurate, full profiles, produced a large positive impact. A pair of hurricanes was found in the NR and their evolution explored in forecast experiments. The landfall forecast was dramatically improved by the simulated DWL observations. Analysis of the hurricane episodes indicates that the entire profile matters, as does the swath, and upper level lidar data matter more than lower level measurements. Observations taken from a single line of sight did not help much. Bob cautioned that the forecast of the evolution of every hurricane might not be improved so dramatically every time. Extra-tropical cyclones have been investigated as well, and the results indicate the value of wind observations increases as the scales become smaller. The impact of AIRS (alone) implies significant impact for this instrument. AIRS and DWL together appear to provide additive impacts. Michael Kavaya inquired what fraction of the total foreseeable OSSE work is complete. Bob did not specify a figure, acknowledging that the results to date are just a start. However, these establish that DWL is needed, even though the quality of the simulated data exceed what is to be expected from a first mission, for which he commends Europe and Japan for pursuing. Wayman Baker wondered whether AIRS would have had a similar impact on the hurricane track forecast. Bob replied that the result was expected by August 2002 (Editorís note: the result was found to be ìno.î) Tom Wilkerson noted that the advantage of DWL (alone) over AIRS (alone) was a fraction of a day, perhaps 9 hours. Bob held that this represents a significant improvement in a situation for which high accuracy has already been realized.
Dave Emmitt (Simpson Weather Associates) spoke on the simulation of expected data products from a hybrid DWL. Support has come from the IPO for hybrid DWL studies. Dave emphasized that this is largely an ìend gameî study, geared toward meeting the objective wind data requirements (for this reason, cross-track sounding is required.) A convenient starting point are the ìreference systemsî developed with IPO and GTWS support, since these provided an idea of what would be needed to meet threshold requirements using the direct detection (DD) and coherent lidars. Choosing a nominal altitude for the hybrid sensor at 400 km, Dave pointed out that only certain requirements are relaxed when going from coherent DWL to hybrid - but these are important: reduced optics, scanner, and associated power. Similarly, for a nominal DD (i.e., double edge filter at 355 nm), the power and optical diameter are reduced when designing for molecular but not aerosol. Since ìpleasant surprisesî sometimes occur in the form of enhanced aerosols but not in molecular density, the DD portion is not driven to cover the lower troposphere. The concept instrument has (rough) requirements of 0.5 J and 0.5 m optics. The assembled hybrid would be ten times smaller than the ìstand-aloneî DD and 70 times smaller than the corresponding CD. It would have one vertical coverage hole at ~ 4 km, which Dave argues might be addresses using an ìenhancedî CD mode, or use of an advanced DD approach, e.g., photon recycling. The next planned activities include conducting the JPL hybrid cost study, working to develop a hybrid technology roadmap with LaRC and GSFC, and performing scaling analysis of TODWL results to help clarify (in part) what the hybrid capabilities might be. Bruce Gentry commented that while the DD portion focuses on molecular signal, it does not preclude aerosol. He also cautioned that the Daveís efficiency figures might be a catchall to account for several factors including light recycling.
Dave Bowdle (University of Alabama at Huntsville) presented recent results of backscatter and velocity measurements from simulated river surfaces using the NASA MSFC pulsed 2-micron Doppler lidar. Measurements using MACAWS (10 micron) and later pulsed 2 micron lasers directed at the Tennessee River had led to inconclusive interpretation of the signal at glancing incidence angles, inspiring the current water slide experiments. Among the numerous factors to consider is wavelength, incidence angle, spot size, distribution of waves on water surface, penetration depth (small for 10.6 and 2 micron). Surfactants play a role, and there are multiple velocities to consider - waves, eddies, and bulk motion, plus near surface aerosol system. Dave showed data from June 5, 2002, obtained with a pulsed laser exhibiting three peaks - water, air, and possibly a hard target (at 0 m/s). These returns are not calibrated or converted to absolute backscatter. Dave listed several advantages to using CW lasers instead of pulsed systems: theyíre small, experiments can be done inside and avoid ambient aerosol extinction, and the small footprint allows for examining single facets, so that a physical separation can be achieved. Using a tunable 9-11 micron CW lidar will facilitate composition studies. Finally, Dave noted the need to improve almost all equipment: lidar, target, and water slide. Lidar improvements include hard target calibration, and the need to simulate a 100% absorbing target. A closed-cycle water supply will enable him to get away from thin flow. Dave was asked to identify the source of the ìhard targetî signal. He replied that this was inconclusive; it is not necessarily from the slide, for example, it could be reflectance from the water surface to the top of the slide. It was suggested that a standing wave might be responsible for the hard target signature. However, Dave maintained, that edge and gate effects were detected, and there were waves, but not standing waves.
Dave Emmitt reported on the Status of the TODWL Spring 2002 checkout flight, conducted as part of the IPO-supported Cal/Val study. He summarized the flights made, noting the MSFC on surface returns motivated the TODWL experiments by raising concern that using surface returns to calibrate the lidar returns from space may be more problematic than had been assumed. The goal is to be able to differentiate returns from the water surface and the near-surface aerosol layer. Chirping is being used to tell whether a hard or diffuse target is being encountered. For the initial TODWL flights in the Spring of 2002, the primary objectives were: to check out operation of a 2-micron coherent DWL and the semi-hemispherical scanning while airborne. The data also provide the means to investigate how TODWL (and similar instruments) might be used to provide cal/val for scatterometers and RadarSat, and an opportunity to add to the database of aerosol studies. Details of the lidar and its aircraft accommodations were reviewed. Flights made to Boulder, CO with ancillary measurements by NOAA/ETL and CTI were described, as were marine flights near Monterey, CA. Although he cautioned that none of the data are fully analyzed yet, he expects to post results on his website: www.swa.com/todwl. Dave listed several upcoming research opportunities with the TODWL, including new GroundWinds intercomparisons, Pre-THORPEX, NAST flight over Hawaii and he offered TODWL as an additional resource for ADM cal/val. He closed with a proposed research plan for ground-based experiments with the surface version of the TODWL. Dave Bowdle commented on the TODWL rapidly sampling a river surface flow, compared to a technician taking hours to do single river-flow sounding in situ. Dave Emmitt acknowledged this but expressed little optimism for river flows being routinely measured via lidar.
Dave Emmitt proceeded to discuss an investigation of the marine LAS (layer adjacent to surface) with an airborne Doppler wind lidar. The LAS contains transient aerosols, and the challenge is to be able to sort out the motion of these from vertical wave motions and surface currents. Data obtained on March 2, 2002 with the TODWL making 30 degree scanning featured double-peaked spectra - one for water and one for LAS. The spectral peaks are especially clearly revealed using a sliding range gate approach. Dave explained that with a hard target he can range with much better precision than delimited by pulse length, due to the sharp change in intensity. Hence, he is confident that it is possible to discriminate between water and air returns. Wave data were available from a NOAA Buoy. Time series analysis of LOS velocities from TODWL, subjected to low pass filtering, indicated comparable (200 m) swell wavelength (assuming that consecutive peaks in the TODWL return are the result of maxima from a crest/trough pair. Dave was asked about the quality of weather prediction around Monterrey, where these flights were made. He stated that it is poor; the army canít predict winds confidently for burning, and hence wants TODWL winds to support Naval Postgraduate Schoolís MM5 output during controlled burning.
Dave Emmitt concluded his series of talks by discussing the nadir angle dependence of water surface return at 2 microns using TODWL. Data from a space-based DWL would benefit from using surface returns to calibrate each shot. Therefore, a model of how signal should fall off as function of angle for different wind regimes has been developed. Using TODWL, only one order of magnitude decrease between 0 and 30 degrees is found, but the theory (model) predicts three orders of magnitude. TODWL data were saturating the detector close to nadir, though, and Dave speculates that had an attenuator been applied, the observations would come closer to theory, but only by one order of magnitude. Thus, this problem remains not solved, Dave expect to share more with the Working Group as progress is made. Several suggestions and questions were made about the interpretation of this work within the framework of the earlier efforts by Tratt and Menzies.
Christian Werner (DLR) presented results of airborne Doppler lidar measurements of the ocean and river surface returns. He described the CO2 TEA laser and lidar system, and described how it successfully measured tropospheric winds during several campaigns (including a jet stream near 9 km confirmed by VHF wind profiler observations at Lindenberg and local model analysis.). With wind sensing established, Christian has turned to examining airborne measurements of oceans and rivers. The lidar was able to detect surface flow as the beam transected a river near Marseille. Moreover, it appears possible to discriminate between wind and water surface (one can see signatures of spray and breaking waves) over the sea. Christian proposes for the analysis of water surface returns, that all of the raw data should be stored and analyzed off-line.
Mike Hardesty discussed new airborne turbulence measurements, the fruit of the recent IHOP campaign (May/June 2002.) Mike had reconfigured ETLís HRDL instrument for operation on a DLR Falcon aircraft to look at turbulence in vertical (nadir) wind field. The objective was to try to use this with DIAL from DLR to try and measure water vapor flux (never done this way). It was a character building experience. The instrument was removed from a sea container and fitted to work in a Falcon aircraft within one week. Despite the limited time and space with which to work, it was installed and collected data. A turning mirror was used to adjust for the Falcon pitch angle and to keep the beam at nadir. Flights were made over surface flux sites, covering the same area several times. Preliminary time-height sections showed up and downdrafts in the turbulent boundary layer (BL), topping out at just meters above the surface. Computing the mean of these motions, they found 10 cm/s or less using block averaging. Velocity variance increased from 50 to 150 m, then decreased above. These features and the skewness profile are qualitatively encouraging. The DLR DIAL data contain variations that seem to correlate with ETLís DWL observations. In summary, Mike is confident that theyíve demonstrated measurement capability in BL. Next steps include cross-correlating vertical motion and water vapor density. They have not yet looked at horizontal data, but they may try that to compute water vapor flux along that direction.
Dave Mordaunt (TRW) briefed the Working Group on the status of TRWís solid state laser capabilities for wind lidar. TRW has done considerable work with solid state lasers and 355 nm development, including diode pumped lasers for numerous applications. They have a 1-micron system that when tripled might deliver 0.33 J at 355. For ZEPHYR, the UV output exceeded 300 mJ. They found good beam quality at 1 micron, and found the UV beam uniform in near field and diffraction limited in the far field. The TRW work at higher PRF confirms this as viable means of achieving the desired average power, though Mordaunt acknowledged that it may not serve the peak-pulse desire of the DWL community. Dave described TRWís Nd:YAG phase conjugate MOPA and noted that it has achieved high pulse energy and excellent beam quality. He also discussed high power laser applications (such as lithograpghy and machining), and lasers used for precision monitoring. Dave also talked about packaging some of these systems for environments outside of the laboratory, and showed compact and rugged hardware built for military applications. He noted that using conductive cooling will require intensive development for space-based diode pumped lasers, noting the high energy/area to be dissipated, and the relevance to lifetime of diodes, especially for three to five year missions. In conclusion, Dave thinks TRW could produce a reasonable sized optical head, and a tripled YAG suitable for space missions. He stated that critical technology issues have already been demonstrated. Paul Hays remarked that with respect to the issue of diode lifetime, Dave (and others) need to maintain close contact with the manufacturers, who are unlikely to abandon their bread and butter telecommunications industry products to pursue a more limited market elsewhere. He also asked about the wallplug efficiency of the instruments discussed. Dave noted that one achieved 7% without cooling, and somewhat less when cooled. Conductive cooling is expected to help in this regard. Finally, Dave was queried about problems of long-term conversion to the UV. He acknowledged the testing over the hundreds or thousands of hours it might take to affect UV coatings has not been done, but is worth an investigation, since UV coatings are typically more susceptible to damage than those in the IR.
Farzin Amzajerdian (NASA/LaRC) reviewed the status of the advanced lidar technology project (ATLP) established by NASA to implement lidar risk reduction. For ALTP, GSFC will concentrate on 1-micron technology (and its harmonics) and LaRC will address 2-micron lasers and 792-nm lasers. The goals are not limited to wind ozone, CO2, and other species are to be sensed. The ATLP charter is to develop technology for future missions, functioning as an end-to-end lidar team. There is to be collaboration between government, academia, and industry, in order to produce validated technology serving a range of lidar applications relevant to NASA. For 2-micron technology an end-to-end approach that considers laser transmitter, scanner, auto alignment, lightweight optics, and a low noise receiver, is being pursued. For UV wavelength conversion, accomplishments of the past fiscal year include developing a 150 mJ, 320-nm lidar with 10% efficiency, and establishing a collaboration with Sandia National Labs to push further (toward a goal of 500 mJ.) Detector development for 2-micron emphasizes CO2 DIAL, for which even lower noise is required than for DWL. Elaborate detector characterization equipment is now in place. For the 2-micron laser transmitter, the goal is a conductively cooled, 1-J lidar operating at 10 Hz and 5% wallplug. Double-pulse operation is being explored for increased efficiency. Already, 600 mJ has been achieved with double pulse and 2% wall plug. Finally, Farzin discussed the laser diode tasks. LaRC plans to secure 792 nm diode pumps and GSFC 808 nm diodes. This is acknowledged as the most critical issue in terms of instrument lifetime and reliability, especially 792 nm. ATLP wants to establish working relationships with manufacturers to advance the theory of laser diodes, and to establish database. A listener agreed with Farzinís assessment of the criticality of diodes, lamenting that 808-nm commercial products have low lifetimes, and they are lower still in arrays. Those used in telecommunications industry exhibit life times orders of magnitude higher - because there is a standard to be met. Farzin replied that for this reason NASA wants a multi-year program with industry involvement, and perhaps to pool resources with DoD.
Farzin Amzajerdian proceeded to discuss the evelopment of mobile validation lidar facility at NASA/LaRC, the scope and purpose of which he reviewed from the Key West meeting. Goals are to extend coverage to higher altitudes, test components, and eventually serve as a CAL/VAL facility for space-based instruments. He showed a photo of the 48-foot trailer, with most components now integrated. First sample winds for the BL were presented, and show evidence of Ekman spiraling.
Rex Fleming (UCAR) discussed the concept of a windtracer on commercial aircraft. With respect to Homeland Security, Rex argued that we need to develop the capacity for mitigating the effects of environmental terrorism, both by minimizing loss of life and by reducing economic disruption by maintaining normal activities as much as possible, and resuming them as quickly as we can. Thus, there are both short and long term responses to be developed. Central requirements are for plume models and accurate wind data for mesoscale models to drive them. Satellites donít have the vertical or horizontal resolution to provide the requisite mesoscale observations. Ultimately, Rex recommends a composite system featuring ACARS, conventional satellites, and satellite DWL. Rex was asked whether ACARS and aircraft DWL should be depended upon, since aircraft may be grounded in the event of an attack. He replied no, since aircraft data obtained prior to grounding would be critical. He also expressed the opinion that there will be great reluctance to shut down (other than locally) in the face of terrorism. Even for an extreme event shutting down commercial air traffic over a larger area, Rex proposes that winds be measured using UAVs. Rex indicated that an ìaverageî airport serving an ACARS-equipped plane will provide 48 wind profiles per day, while even a small airport may generate 3-4 daily. Rex also argued for rapid deployment of automated wind profilers at every port to counter the threat of a cargo container nuclear device. If a device is detonated, weíd need to rely on wind profiles from commercial aircraft, augmented by satellite DWL (needed for night coverage when aircraft flights are fewer.) He recommends:
Questioned about the cost of aircraft modification, he noted that GPS winds from aircraft are already developed and the modifications fairly low cost. For an actual DWL, the first aircraft modifications would be costlier. A commentator noted that the variety of terrorist events relates to a variety of scales. An anthrax release in a city would correspond to the microscale; to achieve Rexís goal, each city might need ground-based DWL or profiler. The value of ìupstream dataî was mentioned, and Rex acknowledged its importance and maintains that it would be provided by Pacific carriers and later from satellite DWL. Bob Atlas noted that by the time the Observations System is complete, NASA DAO and others will be running models at 20 km resolution.
Dave Emmitt reported on ìBetween Flightî Programs for TODWL, since ONR has encouraged use of assets of TODWL to be used between flights provided no conflicts. A trailer has been purchased to house the lidar, and CTI contracted to integrate the hardware. The next TODWL flights probably will occur in February 2003, so rapid integration and breakdown capability are needed. Dave showed how the TODWL scanner is fitted to the trailer. Planned operations include: study of forest canopy energy exchange, cloud dynamics in the boundary layer (BL), engineering checkout exercises for VALIDAR and GLOW in collaboration with NASA, investigations of marine BL and coastal circulations (sea breeze development), and support for controlled burning by the US Army. Several upgrades are being considered, among them: new conductively cooled transceivers, a new data system, and a cockpit display of DWL data. Dave hopes to have the first ground-based data results by the next meeting.
Geary Schwemmer (NASA/GSFC) presented HARLIE Wind Profile Measurements from the HARGLO-2 campaign. He acknowledged that processing data from older campaigns has been delayed somewhat by participation in new field exercises, including IHOP. However, a lot still has been done, especially with regard to wind retrieval algorithms developed by Tom Wilkerson and USU students. Geary reviewed the objectives, measurements, and post-processing procedure (30-min averages re-gridded to 200 m vertical centered on the odd hundred m level) from the Key West meeting. For the preliminary intercomparison matrix, no data filtering was permitting, allowing HARLIE retrievals, both manual and automated, to be assessed statistically and the optimal adjustments made later. Data availability predictable peaked in the BL and from cloud layers - stratus and cirrus. Scatter plots of HARLIE winds (manual retrieval) versus all sondes contained 248 points with a fair amount of scatter, with a bias of -2.2. m/s and 20 degrees. Comparing radiosonde, HARLIE (manual), and GLOW data, there is evidence that speed and direction offsets between sonde and HARLIE at upper gates may be caused by aircraft contrails. Plans are to finish HARGLO-2 analysis, then IHOP and HARGLO- 3 analysis. However, more campaigns are in the works, including the ìmarriageî of HARLIE and GLOW (1064 nm and 355 nm). Geary was asked to compare HARLIE wind measurements with the methodology used by Ed Eloranta for a number of years. He noted that the approaches are similar, in that both use azimuthal scanning to infer non-Doppler winds, but Eloranta actually scans volumetrically, in three dimensions. Geary expects to remove the limitation of single-axis scanning for HARLIE.
Michael Kavaya (NASA/LaRC) outlined a technology plan for the coherent subsystem of a space-based hybrid Doppler wind lidar. The target is a 0.5-J, 12-Hz lidar qualified for a 5-year mission. The technology path starts with the laser transmitter first demonstrated for a ground-based laser, then an airborne sensor, and finally a space-based mission. A question was raised about the proposed time line. Michael replied that this depends entirely on how open oneís checkbook is. In the best of all worlds, perhaps preparations for a launch in 2006 or 2007 is reasonable - but not without money. Wayman Baker commented that it is too late to be included in NOAA and NASA budgets before FY05. Although the STP might provide launch, a year of operations, and ground processing, the challenge is to get a program in synch with this as soon as possible. Michael resumed by considering the scanner for which he baselines Si wedge technology, since it is validated for decades. Alternatively, a rotating telescope scanner (and maybe share with DD subsystem) could be considered. Pure electro optic scanning has the promise of being free of moving parts; if it can be demonstrated with a ground system would like to consider adapting it for space. A LO Laser has been developed (by JPL) and implemented (by CTI) as a baseline; improvements are expected. Michael notes that increasing photon efficiency is difficult to fund. Despite that, increasing efficiency achieves the same gains as having a bigger laser or mirror. He identified auto-alignment technology as the least well funded and developed of these various elements. Another challenge would be a dual-wavelength telescope/scanner. Much theory as well as initial fabrication for ground-based version is needed. In conclusion, a hybrid approach may be the way to go. The hybrid technology roadmap has similar qualitative look (milestones) as the go-it alone versions. Michael pointed out that spacecraft considerations were prominent in the CD and DD for ISAL exercises, and would be in the corresponding efforts for the hybrid.
Bruce Gentry similarly presented a technology roadmap for a direct detection doppler lidar subsystem. At the cartoon level, he noted, the plans are somewhat similar; both require a high spectral resolution solid state laser. A critical need is to validate the technology components and how they interact with the atmosphere, by surface and airborne work. For DD, other key elements are high-resolution optical filters and photon-counting detectors at 355 and 1063 nm. Novel large aperture telescopes and scanning optics are also requisite. In its basic form, the NdYAG is well along. Development needs to focus on efficiency, lifetime, and conductive cooling. The function of high-resolution optical filters is clear for DD. As a baseline approach, Bruce recommends Fabry-Perot, while noting that Mach-Zehnders and Fizeaus are being considered in Europe for ADM. The challenge for DD detectors is the need to count only tens to hundreds of photons per gate in the upper atmosphere; they need to be low noise. A large dynamic range is needed to accommodate ground and cloud returns. As a baseline approach, Bruce recommends efforts to improve PMT photo cathode efficiency, and reduce read-noise for accumulating CCDs, and to consider APDs as well. Bruce closed by noting how ground-based systems such as GLOW, GroundWinds, and others are providing a sense of end-to-end efficiency that is useful for guiding these deliberations. Dave Emmitt commented that autonomous operation should be considered on the roadmap. Bruce was asked for a rough estimate of resource (power) requirements? Referring to Dave Emmittís presentation, Bruce indicated 200 mJ at 355 nm.
Evan Chicklis discussed Tactical Laser Development efforts from the military contractorís perspective; although not directly related to wind sensing, there may be some common ground and lessons to be learned. At BAE (formerly Sanders) in NH, efforts at building 1 and 2 micron lasers for IR counter measures, fire control and target ID, and prototype remote sensing, have provided insight to engineering issues associated with making lasers, including diode reliability. The key tactical requirements are low cost, compactness, no consumables, and reliability; even in big programs, the laser should not cost too much. BAEís progress and accomplishments include the so-called SNIPER laser for Advanced Targeting Pod, which was delivered in 16 months start to finish. Heat (thermal gradients to be precise) had to be minimized. Nd:YAG slabs were connected to bench for conductive cooling. Whiles hardware can be developed to provide both high PRF and high energy (pulses) via a combination of diode pumping and conductive cooling, Evan regards pulse energy as the enemy. If you can get (and use) the same average power at higher PRF and lower energy, he recommends taking that approach. An IR Counter Measure (IRCM) lasers have been developed so that a helicopter emitting a heat signal and thus becoming a target can effectively mitigate that risk. IRCMs defeat heat-seeking missiles by modulating the target, since the missiles seek a d.c. signal, modulation effectively confuses them. Evan listed a number of engineering issues for military lasers which may be similar (or worse) for DWL: optical mounts (and rigidity thereof), contamination control, temperature control of critical elements, heat removal, damage testing of all coatings, active testing of all optical elements. He singled out the importance of diode testing, and cautioned that the military-qualified laser diode that BAE has acquired resulted from lots of close work with manufacturers. Diode delaminating tendency had to be overcome, since an Au-In interface degraded over time as a result of Au diffusion. In summary, he noted that 3 x 10 8 shot-lifetime specification has been met. Long-term reliability is being approached and diode cost reduced by a factor of three by working with the vendor and buying in bulk.
Michael Kavaya next talked about ìCoherent Lidar: Factors of Two Among Friends.î A long-standing issue in the coherent DWL community is that SNR(exp) < SNR (theory). In essence, coherent lidar measurements of backscatter (beta) do not agree with other sensors. Michael outlined the papers that have described the discrepancy, and summarized their major points. One hypothesis is that the disparity results from differences in target calibration methodology. Another points to the underestimation the effect of refractive turbulence, which Michael is inclined to reject because it should not give a factor of two but a continuum of values. For the same reason he discounts the effects of optical aberrations. Having compiled a matrix of observations and ìsuspect compatibilityî, he deems the most likely suspects to be the ìopposition effectî and target calibration (Tratt and Haner). Ultimately, Michael is seeking opinions and advice to sort this question out. He noted the comforting fact that the issue does not affect predictions for space - as long as scaling predictions for space are based on the more conservative experimental results. However, theory should be developed to try to reconcile from first principles, and additional experimental work on the opposition effect are needed, too. Jim Hawley wondered whether polarization effects could account for the difference. Michael replied that heíd shown an abbreviated list to meet time constraints, but he is fairly confident the prime suspects have been identified.
Sid Wood (SWA) reported on the status of the Doppler lidar smulation model (DLSM). Sid has completed the following since the Key West Meeting: Set up server and firewall; ported 1995 DLSM into PC environment; developed DLSM user registry code; developed a new C language graphics and toolkit; re-coded LSM and shot coverage into Fortran 90 and set up on server; added two DD algorithms to the model and a CD algorithm support for ECMWF and FVGCM atmospheres. Alpha testing of the model is 95% complete. It will be released to the community for Beta testing soon. Sid presented examples of simulated data extractions, emphasizing the ease of use and flexibility of the tools. Near-term activities are to finish vertical slicing of LOS product, compile communityís input from Beta testing and complete help files and userís manual. Future tasks planned include improving the orbit model, integrating the coherent and direct detection algorithms, developing access to FASCODE from the DLSM, and developing an error model for pointing and navigation. Sid was asked to estimate when this would be done. He replied that release for beta testing should be in August 2002. He was also asked whether the source code is in the public domain, and he affirmed that it is. One can ask to see, for example, how direct detection is being handled in the code. A follow-up question was whether the system is capable of having user algorithms for different configurations? Sid indicated that the intention is to serve as a community-based model. Anyone can come to the table with a request, have it reviewed, and integrated (for all to see and use), if itís appropriate.
Bruce Gentry (NASA/GSFC) gave an update on GLOW-based on tropospheric wind observations made during the HARGLO-2 and IHOP campaigns. GLOW, as a ground-based DD DWL provided wind measurements. This talk emphasized HARGLO-2 (November 2001) data. Bruce presented statistically determined error bars, increasing with altitude as the signal level dropped, and also random wind errors predicted by shot noise. He argued that the best way to intercompare wind measurements such as these it is to bin them by the error estimate, so as not to dominate with shot noise error. Bruce briefly described GLOWís participation in IHOP, noting that at the Homestead profiling site between May 14 and June 25, about 270 hours of data were collected, usually at four azimuths, but occasionally as RHI or PPI sections. He discussed system parameters and choices that were explored: reduced aperture and laser energy (to see whether this cut down background noise), and the high PRF, reduced peak pulse energy approach. A quick look indicates that data are OK; given the volume and quality of coordinated data, Bruce expects to be able to do a lot of science. Asked whether heíd compared these data with wind profiler observations from IHOP, Bruce replied that there has not been time yet to do so quantitatively. He was also asked whether any observations from frontal passages had been collected, and he affirmed that such cases had been sampled.
Jim Ryan (University of New Hampshire) welcomed the working group to GroundWinds and outlined tour logistics for the evening. He discussed the collaboration between UNH, MAC, The Mount Washington Observatory (MWO) and later the University of Hawaii. The goal, he emphasized is to answer the questions by stages, will DD work? and will it work in space? He reviewed GroundWinds-New Hampshire (GWNH), showed how the site and facility have evolved, and discussed the variety of conditions, both clear and aerosol; data were collected soon after operations began. GWNH still requires on-site operators and aircraft clearance for operation. GroundWinds-Hawaii (GWHI) was built on the lessons learned from the early operation of GWNH, and in turn, these lessons were refined and fed back to upgrade GWNH. Specific objectives for GWHI are to demonstrate clear air (low aerosol) capability, remote operation, provide a data point for mid-Pacific prediction models, and geophysical and meteorological data to support Mauna Kea astronomy operations. In response to a question he indicated that GWHI operates at a fixed (45 degrees) elevation. Jim reported that retrofitting at NH was completed in July 2002, and GWNH is supporting AIRMAP with wind and aerosol data. He noted that AIRMAP is the first full-up science data provision activity for GWNH. Jim was asked whether GWNH was operated for AIRMAP 1 in 2001. He responded that it had been, but noted that the biggest feedback was for NH and learning how to upgrade the system.
Carl Nardell (Michigan Aerospace Corporation - MAC) gave a GroundWinds project summary and status report. He reviewed the key GroundWinds (GW) technologies (CLIO, streaking CCD camera as detector, fiber-coupled FPIs, signal processing algorithms, and light recycling), the auto-alignment system, long pulse Nd:YAG laser, and the control and safety system. He presented a GW technology roadmap, noting the evolution from what was in hand originally (1998), to the developments culminating in the first wind data from GWNH-2 (July 2002), and speculating on what might come to pass by 2004 for a balloon-borne demonstration. A matrix of design considerations was used to illustrate what items become critical to progress from lab to ground to balloon to space. An articulating telescope and auto-alignment system are considered essential. Data processing on board is anticipated to reduce downlink (although this might not be favored the science community). Carl maintains that GroundWinds has demonstrated most of the critical requirements, except for laser transmitter efficiency/lifetime. He ranks light recycling as GroundWinds most significant accomplishment to date, and stated that it will perform near the limits predicted by theory when strict manufacturing tolerances are achieved in subsequent GroundWinds systems. Carl stressed that GWHI, operating at 355 nm in a location where aerosols are scarce has demonstrated improved efficiency based on lessons learned from GWNH. He pointed to a record of acceleration of completion (to meaningful data products) of each generation of GroundWinds, noting that the GW program has been helped by others, notably MOADS for USAF, and via Navy and NASA SBIRs. MACís assessment in terms of NASAís TRLís: theyíve advanced some items to level 5 (from about 2, with 6 indicating space-readiness minimum.) In the future, Carl expects that MAC will continue to work on efficiency of light recyclers, foremost by applying assembly to achieve specifications to optimize throughput. Improved machining of CLIO elements and the used of photo counting array detectors will be pursued to increase throughput and efficiency, too. A balloon-borne mission is being considered as a means of elevating TRLís closer to what is required for space. Carl was asked about a bias arising from the GroundWinds spot pattern, which had been discussed at other working group meetings. He replied that has been resolved for earlier data sets by post-processing. Since the bias arose due to was instability in the laser and etalon, it has subsequently been removed by hardware adjustments.
Michael Dehring (MAC) provided an overview of the GroundWinds - second-generation system performance at New Hampshire and Hawaii system performance. The thrust of the presentation was the efforts put forth to understand the transmission characteristics of the instruments, which began with examination of all optical components that were reasonable to test. The fiber coupler (FC) transmission testing revealed losses of 8-16%. Subsequent investigation indicated that these arose because (1) hole location was not precisely symmetrical enough, and (2) the fibers were too small, allowing them to fall out of position (to the side of the hole). The problems are being corrected. Based on the performance of the individual optical components, GWNH-2 was predicted to achieve 0.45 transmission; the actual figure is 0.39. Likewise, GWHI was predicted to reach 0.46, and in fact, attained 0.45. Michael discussed how etalon control validation - via real time processing - to determine post motion per volt input - is done in the field. He showed new GroundWinds wind data, fringe images, error statistics and how they improved, both for GWHI and GWNH-2. He stated in summary that both GW lidars can reliably collect wind data now, with no large systematic offsets identifiable. Both have been fairly well characterized, and the major improvements anticipated for the future are via higher recycling efficiency. Dave Bowdle asked whether any more independent wind data comparisons have been made. Michael answered not yet. Chris Werner commented that one often makes measurements of hard targets to get statistics, and wondered if this has been done with GroundWinds. Michael answered that it had not, at least not yet. Rex Fleming advised that ACARS data be used in place of radiosondes for validating GWHI observations. Mike agreed that this idea has merit.
Jim Yoe (NOAA/NESDIS) discussed results of wind intercomparisons from the GroundWinds 2000 demonstration campaign. He stressed that the best way to make intercomparisons was to apply a statistical approach to a temporally extensive set of observations, concentrating on data obtained when all instruments are close to optimal performance (as qualitatively defined by the respective operating teams) and based on the ìbest and finalî re-processed data gridded to common time and height intervals. By this means one can avoid drawing conclusions that are too optimistic or pessimistic based on single ìtypicalî examples. Using wind data for a period of ~ 8 hours for September 28-29 2000, histograms of LOSH differences for GroundWinds and GLOW, GroundWinds and the ETL mini-MOPA, and GroundWinds and radiosondes were compiled over all heights (and coincident times) for which any pair of sensors was reasonably expected (based on signal strengths) to provide ìgoodî wind measurements. The difference histograms for pairs of DWL were Gaussian in shape with small non-zero means, indicating that all were measuring the wind and its temporal evolution. The spread of the GroundWinds-GLOW exceeded that of the GroundWinds-MOPA, as would be expected based on the weaker echoes and increased uncertainties at greater altitudes. The DWL-sonde comparisons were skewed, indicating the inability of even frequently-interval (90 min) sonde releases to sample the evolution of the atmosphere as effectively as the DWLs, and the spatial separation of the measurements resulting from the balloonís trajectory. This has implications not only for ground-based intercomprarison campaigns but also for the eventual validation of a space-based DWL. Jim was questioned about the mean differences and replied that a height-reporting offset of 0.5 km for GLOW may be part of the cause. Instrumental offsets in the newer lidars (GroundWinds and GLOW) may also be at the root of these. He was asked whether another campaign at GWNH would be mounted to take into account the lessons learned and the improvements in hardware. Jim replied that this would be useful, in his view, if funding can be secured.
Tours, Subcommittee Discussions
Rex Fleming summarized the sub-committee discussion on specifying upper air observing system components to support the Homeland Security mission. The group had agreed that commercial aircraft (CA) should be used to provide temperature, moisture, and wind data from in situ sensors. For critical tactical situations, the system should be enhanced by DWL and perhaps DIAL on CA and UAVs. Arguing that there is a national need to augment this with space-based wind measurements, he urged the Working Group to support this plan. Bob Brown argued that for the West Coast scatterometer data through the depth of the boundary layer should be included in this system. Several questions were raised about UAVs and instruments on them requiring radiation hardening. Rex replied that this needed to be determined. For CA DWL, he believes that they would require one year to certify, with ten systems delivered in the second year, 20 in the third, and doubling delivery each year until a plateau of 1000 CA is achieved. For ìspecial aircraftî bearing DWL, a fleet of about 20 is envisioned. The subcommittee could not agree adding DIAL and Quantum Cascade (QC) lasers for species identification. While this was agreed to be useful and suitable (small enough) for CA deployment, and that communications and data formats for DWL could easily be adapted, the Working Group has no expertise in this area. An additional PI would need to be recruited, and the risk of diluting the main message (winds) would be incurred. Bob Brown questioned how the data will actually be used. Bob thinks there should be a proposal to use surface wind data from scatterometers or SARs over all the oceans. Dave Bowdle asked whether technology is as ready as implied, - for production, integration, and data products, or whether a rapid technology ramp-up is required. Rex maintained that the ramp-up is done for in situ CA winds; for DWL a moderate ramp-up for special planes and somewhat more for UAVs is realistic. J. Hawley commented that key infrastructure (ground stations, etc.) will be stressed when a full complement of planes is transmitting in real time.
Ron Schwiesow (Ball Aerospace) described the space-qualified hardware for the CALIPSO Lidar, to provide a picture of actual space hardware that might bear on future DWL efforts; even though CALIPSO is a straight nadir-viewing cloud aerosol backscatter lidar it might serve in some sense as a pathfinder for DWL. CALIPSOís Nd:YAG laser transmitter is designed to provide 2 x 109 shots at 20 Hz PRF, and the system features 1-m optics and a telescope design borrowed in large measure from GLAS. Auxiliary instruments for the mission include a wide-angle camera to provide context and an imaging IR Radiometer provided by CNES. The receiver module is integrated on single monolithic optical bench. Ron noted that NASA has become cautious with respect to lasers and emphasized the extensive risk reduction for lifetime, contamination control, KTP doubler operation, and misalignment tolerance for CALIPSO. There are two completely redundant lasers to make sure that the program would advance - each designed to meet the entire three-year lifetime. Likewise redundant beam expanders were included. CALIPSO will use both the fundamental and doubled frequencies. Aspects of CALIPSO that depart from the needs of a DWL include the 25 pm linewidth (at 532 nm), and the fact that the laser is not seeded, simply an MO with no amplifiers. The ìsandwichî filter etalon for 532 nm is simply a solid space between two solid mirrors, representing the notion that it is better to make it to an exacting specification rather than trying to make it tunable over a full free spectral range. It has passed thermal and mechanical qualification and exhibits peak transmission of at least 0.81. Bireflectance cross talk is very low. Ron also emphasized practical aspects of preparation for final integration and testing. A composite model qualified with masses in place has been fabricated, and a payload simulator to ready the US instrument for the French bus. Launch (CALIPSO and CLOUDSAT) is slated for April 2004 on a Delta II 7420 vehicle equipped with a 3-m standard fairing. A question was posed about heat removal. Ron indicated that the lasers are the principle source, and removal is accomplished via conduction. Information about the source and amount of funding was requested. Ron replied $20 M was received from the ESSP (second round). He also indicated that Ball did not contract the lasers with Fibretek, - the lasers were provided by the U.S. Government. The status of Thermal Vac testing was questioned, too. Laser testing is incomplete, but the etalon has been completely thermal vac tested. Finally, the source of the etalon was sought; it was manufactured by a small shop on the Isle of Mann.
Dave Bowdle gave an update on the Regional Atmospheric Profiling Center for Discovery (RAPCD) plans and status. Chimneys and siding are now in place. The platform for the lidar dome is ready, but the dome is not in place. A hemispheric-scanning O3 lidar will be housed in the dome. Sites for a 2-micron pulsed DWL, an FTIR and other instruments were shown. He emphasized that building/housing construction is progressing well. The basic structure is complete, as are all of the required roof penetrations. Casework installation is almost complete, and installation of ceilings, utility bars over optical benches, finished flooring walls and doors is pending. Dave expects to occupy the building in August or September 2002 and transfer the DWL and other apparatus soon afterwards. A lightning system protection is tentative, but the need for it is recognized.