Wayman Baker (NOAA/NWS) brought the meeting to order and reviewed the Action Items from the February 2003 meeting in Oxnard, CA. The status of these items is summarized as follows:
Pierre Flamant presented a review and status report on the Atmospheric Dynamics Mission (ADM.) ADM was selected in 1999 for a 2007 mission and is fully funded. Strictly speaking, ADM refers to the mission; ALADDIN is the satellite, and AEOLUS is the DWL instrument name. A "firewall" is maintained between the ADM science Advisory Group (ADMAG) and industry, with ESA management in the middle.
Instrument baseline review will be in July 2003, with PDR in September, and phase C/D to begin in October. For procurement purposes, the instrument is broken into some 50 subsystems for open competition. The majority of the sub-contracts are to be assigned by July 2003. Galileo Avionica has been selected for laser. A Fizeau interferometer has been selected for the Mie channel receiver. Two preliminary contracts apiece are being let for seeder and burst mode diode development studies.
An AEOLUS end-to-end simulator project is to be kicked-off in October 2003. This will be done by industry with ADMAG science guidance. Ground stations are being worked by ESRIN. Pierre also described supporting science studies, including those to optimize ADM data assimilation for NWP, quality control methodology and tools, level two wind processor algorithms, and level two cloud/aerosol products, and maintenance of AEOLUS reference simulator. Data dissemination, post-launch validation, and outreach have not been discussed extensively to date. The next ADMAG meeting is scheduled for mid-October, 2003.
Ingrid Guch asked whether there are any international agreements between ESA and US government agencies for ADM. Pierre replied that there is ADMAG participation from the U.S., and mentioned ongoing discussions about support from US ground stations, for example, Point Barrow, AK.
Oliver Reitebuch discussed DLRís plans for airborne DL campaigns to support design of an airborne ADM demonstrator. Motivation is considerable since the combined (Mie and Rayleigh) receiver will contain a lot of new technology. Also, the ADM optics have been characterized well in the lab, but these studies have not taken into account the complexity of the interaction of the optics and the atmosphere. Preliminary error assessments for simulations use only one (ESA) standard model atmosphere and need to be broadened and extended. There has been no direct detection airborne experience, and clouds, background, and ground returns looking down must be characterized.
Three campaigns planned for 2004 ‚ 2006. The first will be in 2004 at Lindenberg. A UHF (482 MHz) wind profiler will provide the "truth" from the ground to 15 km with 250 m vertical resolution. This system operates "24x7" and already has been proven through three years of intercomparisons with four/day rawinsondes over flat terrain. The 2005 and 2006 Airborne DWL campaigns also will take advantage of Wind Profilers in Europe. These campaigns will use the Falcon aircraft. Previous results from S. Rahm were reviewed, showing strong SNR close to the aircraft and from the PBL, with gaps in between. Comparisons to dropsondes released from the same plane are encouraging when SNR permits wind determination.
Oliver discussed specific parts of the simulator, including laser and receiver, 20 cm telescope, aircraft window, and cooling unit. Aircraft accommodation studies are underway. He also reported on studies of the laser pulse characteristics and Monte Carlo simulations for a double edge filter and shot accumulation in the atmosphere. DLR is trying to include the Brillioun doublet in simulations, using the DWD model with shear and inhomogeniety as input to simulation.
Ingrid Guch discussed current and future DWL activities at NOAA, explaining her role at the Office of Systems Development as Special Projects manager. She outlined NOAA's need for winds and current strategies for meeting these needs, including NOAA's strategic planning process and the recent emphasis on matrix management. She indicated that the upcoming international Earth Observation (EO) Summit bears watching, to see whether wind observations and sensors are emphasized. Ingrid recommended that a DWL FAQs be assembled and posted by the Working Group. Since we have in effect an assessment of where we are and what we need with respect to wind data, but not the means to attain them, she also recommends that we continue to emphasize the roadmap.
Ingrid stressed development of tie-ins to the NOAA Strategic Plan. Finally, the EO Summit, hosted by the DOC and DOE, will represent 30 countries invited at the ministerial level. A Draft Declaration to be presented from the U.S. Group calls for full and open exchange of observations from all platforms with minimum cost and time delay. A 10-year implementation plan is sought. The Summit is very high level, and that the ad hoc working group will be at the NOAA Administratorís level.
Steve Mango reported on the Status of NPOESS. He covered schedule, instruments, NPOESS Preparatory Program (NPP) and emphasizes bridging TERRA, AQUA, and NPP to NPOESS from DMSP/POES. The IPO is progressing toward 2009 Launch (first of six) and also toward individual spacecraft and instruments lasting seven years. Current end-to-end EDR latency estimates 95% of data delivered in 28 minutes or less, with the average under 10.5 minutes. Data rates expected are 12 MBS from the IR sensors, 3 MBS from everything else.
NPOESS will adopt the shared system performance responsibility business model between government and contractors. The Joint Calibration, Verification, and Validation Working Group is an example of the complex but close connections to this approach. End-to-end product development, remains the goal but is not assumed at face value, and heritage emphasis will remain as algorithms are adapted and evolved from the existing base. Wind profiles remain foremost of the 21 unmet EDRs and a target for NPOESS PPPI. Risk reduction activities related to DWL funded by IPO include OSSEs, and hybrid feasibility.
Rex Fleming updated the status of a proposal for Homeland Security for a Nuclear-Biological-Chemical (NBC) mitigation observation system. Three types of time response, corresponding to saving lives out to when ground zero can be returned to economic viability, were taken into consideration for the proposed observing system (OS). Wind data are key because plume models are driven by a mesoscale model lacking mesoscale data input. From categorizing combinations of wind strength and atmospheric stability as well as weapon type, Rex concludes that two types of OS are needed: (1) strategic to cover the whole country before an attack, and (2) tactical that is brought in as a response to an event. The tactical OS would consist of a series of systems in place, on call, stationed with bias toward potential targets. This would include a commercial aircraft wind and temperature sensor, plus water vapor. To provide 4-D coverage around ground zero, Rex foresees the need for ground mobile, UAVs, and manned AC. He argues that Homeland Security will require technology that is (1) proven and (2) can be put in place quickly, and (3) tailored to HS issues. As additional strategic items, he suggests 20 coastal city wind profilers; 20 water vapor DIAL systems, and 20 airborne Doppler wind sensors that could be based on the CTI prototype.
Dave Emmitt reported on the status of the IPO funded Airborne Lidar Experiment. For 2003, the Twin Otter DWL (TODWL) objectives are MM5 validation, to provide NAST under flights, serve as a dry run for THORPEX, and producing winds from cloud returns for comparison to cloud-motion vectors (CMVs). He showed the aircraft and instrument, reviewed characteristics. The lidar can measure sea wave spectra as well as wind components, and will be part of the IPO cal/val suite. The MM5 validation campaign was conducted at the Naval Postgraduate School at 4 km and used for prediction of smoke transport for burning near Fort Ord. For land data, Dave has used a threading approach to discriminate between land peak and true wind in spectrum, based on a direct comparison of vertical radial velocity to check DC offset of VAD.
For NAST (and GIFTS) validation, four hours of ER-2 hours of flight are available, with underflights by TODWL. Though these are not flown at the same height or speed, the CMV compared well to direct TODWL data. General agreement with buoys and scatterometer has been found, too. Organized large structures are being investigated, for example, layers in the range/attenuation through the PBL, and low level jets and shears. Shear causes spectra to smear in a non-Gaussian manner, for which advanced signal processing techniques will be needed. Comparison between models and the actual DWL performance is being done iteratively, with modeling that includes the instrument and the atmosphere. In summary, TODWL's sensitivity is close to what was expected, with water surface returns higher than thought. It achieves accuracy of 0.2 m/s and is becoming more reliable, though not turn key yet. Future activities include WindSat Validation work scheduled for Fall 2003, participation in THORPEX-4, and perhaps a role in a DWL hybrid demonstration with NASA/GSFC.
Bill Heaps presented some thoughts on scanners for DWL. Through his role as a laser advisor for the ISAL, Bill became interested in alternatives to the large baseline telescope and scanners in the GTWS reference designs. The first alternative (for direct detection) might be HOEs, despite large power requirements. Any continuous scanning requires lag angle compensation, and summing shots is complicated. A solution is to have a large mirror scan continuously, but do compensation with smaller elements in the optical train. Alternatively, an off-axis parabola cold be used as scanner AND focusing element. This introduces a new problem for large off-axis angles, in that displacement in the focal plane gets large. Other implementations, such as breaking a rotating parabola into smaller elements, are also possible.
Bruce Gentry presented results of GLOW Wind Measurements from IHOP 2002. IHOP's participating sensors included seven ground-based and aircraft lidars, as well as mesonet and ground-based instruments such as radar. GLOW was deployed for six weeks at the Homestead site and accumulated 210 hours of data over 34 days. Sampling was step-stare at the Cardinal points, with varied elevation. The low energy mode ‚ reduced power-aperture product ‚ was selected to cover the PBL without detector saturation, and it still performed up to 6 km, which permitted sampling of phenomena such as a low-level jet and bore event. Quick-look routines focused on radial velocity data and time-averaged speed and direction profiles. More quantitative analysis with improved receiver calibration, aerosol and cloud effects, analysis of PPI and RHI sets, advanced visualizations, and comparisons with other wind sources are underway in the post-processing time frame. Bruce intends to make data available to the community in NetCDF format later. GLOW is both a science and an instrument technology testbed. For IHOP 2002, he tried a new high-brightness solid state transmitter for wind lidar. Frequency stability was not optimal for the campaign but has since been achieved. Average power is expected to go up by factor of five when PRF is increased to 50 Hz. Currently Bruce is evaluating trades for an AC DWL for IPO, especially to determine whether any of the GLOW (and HARLIE) components can work for this purpose. Bruce looks favorably on ìchopping upî the transition to space (ground demonstrations, Balloon Winds, and ADM.)
Mike Hardesty reported on Combined Doppler/DIAL of studies of moisture transport and mixing during IHOP. The analysis was made possible by a unique DIAL and DWL deployment on a single aircraft, the DLR Falcon. Integrating the DWL into the limited space in the Falcon was problematic, and led to operational challenges, especially for heat rejection. Not only was there added heat from the DWL, the DIAL normally is flown higher, in cooler ambient air. Vertical velocities show several m/s in the lowest 2 km of the PBL. Vertical velocity spectra were calculated to ensure that the variability is atmospheric in origin before calculating fluxes. Can detect Updraft plumes can be detected in DIAL data as well as in velocity data, although the wind structure is more organized. Thus, Mike was able to calculate a flux profile. For horizontal winds, a wedge scanner was used, fixed to 20 degrees off nadir. They were able to observe LLJs, and horizontal winds were successfully compared to dropsondes showing success in removing ground-relative aircraft motion. The Lidar and Dropsonde Flux comparisons have consistent features. Aerosol and vertical wind correlation are being studied now and are likely to be relevant to a space-based lidar.
Dave Emmitt discussed the concept for a Hybrid DWL, with focus on the tropics and tropical cyclones as requirements drivers, such as the transition of African Easterly waves to hurricanes, but which can't be see in temperature, humidity, or cloud fields. Also, local convergence and precipitation are more tightly coupled in tropics, and there are lots of clouds with a bimodal distribution ‚ high cirrus and low stratus. OSSEs show the need for BOTH high and lower tropospheric winds. Mission parameters for the Hybrid call for a 400 km orbit (which lets pulse energy fall by factor of five relative to an NPOESS orbit and also a smaller telescope can be used). Staring and accumulation are critical for the concept instrument. Oliver Reitebuch noted the hurricane emphasis, and commented that this ignores the needs of the Western U.S. as well as Europe. Emmitt replied that the hurricane emphasis was for a demo/science mission prior to an operational mission.
Geary Schwemmer presented scanning strategies for holographic receivers to stimulate discussion on how strategies relate to HOE designs. Significant aspects of scanning include whether the scan angle must be adjustable, and step-stare versus continuous versus targeted observations (which would drive to dual axis scanning.) Single rotating HOEs (reflective or transmission) have been done, are simple and provide an infinite number of azimuths, but difficult to adjust and require momentum compensation. Shared Aperture Diffractive Element (SHADOE) promises lowest inertia and thus is good for step-stare, but may need multiple lasers. Geary noted that the scanning HOE could save 100 kg; and removing scanning by going to a single perspective might save another 100 kg on the GTWS ISAL baseline. The SHADOE approach also saves ~ 100 kg in weight.
Maximum efficiency is achieved at 45 degrees, which is suitable for wind profiling. Questions were asked about the efficiency of HOEs for various wavelengths, and the related problems of absorption and heat loading, which he acknowledged is especially worrisome for 355 nm up to this point.
Ken Miller gave a summary of the DWL Roadmap, which has grown out of the work of the GTWS Executive Steering Committee, drafted as a Lidar Working Group action item, and has been submitted recently for NOAA and NASA management review. The GTWS reference designs were reviewed at a concept level to show how these led to the Hybrid roadmap to take advantage of complementary aspects. Ken recommends an ISAL effort for a hybrid DWL be undertaken to extend Emmittís feasibility study. Roadmap issues include the designated technology development areas. There is a need to explore the sensitivity of the benefits of the data requirements since these drive the technology challenges. Trade studies between direct, coherent and hybrid approaches should incorporate instrument, atmosphere, and spacecraft mechanics, and calibration and validation. The timescale for the roadmap will depend on funding and resource decisions foremost, then technology advancements. Extensive discussions followed. Steve Mango suggested that the roadmap target additional EDRs such as cloud and aerosol properties to tie it to other potential users.
Dave Emmitt addressed the relative benefits of full versus partial wind profiles as an on-going work, laying the basis for new OSSEs. Users state different preferences for low and high level winds. Some older scatterometer studies indicated that the "right" models can couple lower level winds successfully to upper levels in some circumstances. The recent DAO OSSEs indicate the need for full profiles. Asked what to do first if we can't afford everything, Dave indicated that the answer is application (user and customer) dependent. A comment was made that since the scatterometers give ocean surface winds, it might be fruitful to concentrate on upper levels with DWL. Asked about the value of wind versus water vapor profiles, Dave pointed out that the lack of wind information hinders studies of water vapor transport more than the lack of moisture data.
Ivan Dors gave a status report on GroundWinds New Hampshire's (GWNHís) Operations and Science, reviewing system specifications, the measurement principle, and data products (u, v, w wind components, spectra, aerosol/molecular backscatter and optical depth, and temperature.) In the future, GWNH will be used to study Brillioun scattering. Current emphasis is on spectral determination of white noise and velocity turbulence. Questions of optimal measurement and interpretation are now being worked, using a Monte Carlo simulation that includes time series, noise, trend, and turbulence. This permits extracted turbulence parameters to be compared to known inputs. The precision of turbulence estimates is based on the uncertainty of wind estimates, which depends on time averaging. Ivan estimates that the optical white noise and turbulence estimates are produced from 1 s data.
Ivan Dors continued with the GroundWinds Hawaii (GWH) Operations and Science summary. GWH currently has no aerosol channel, as the hardware is being used for the BalloonWinds development. New GWH data are being posted on the Web. Ivan described how decreased uncertainty in radial velocity estimates has been achieved as the system is optimized, with less than 1 m/s at 5 km and less than 4 m/s at 15 km. GWH is being operated by UNH for Mauna Kea support and for THORPEX. Currently, UNH is taking wind data and using CMV data (from a NOAA satellite) in the LAPS model (and eventually MM5.) The lidar was also used successfully for the PTOST campaign, providing the only ground truth. Ivan showed an example of a jet and strong shear at 15 km.
Michael Dehring provided the BalloonWinds Update. The baseline data collection goals were described. Simulations show that the instrument will be sensitive enough to go with very short averages. The baseline mission duration is 4 h, but depending on conditions and batteries, there is hope for as much as 12-16 h of data collection. With 3-4 W at 355 nm and 30 Hz, radial velocity profiles are to be produced at 12 s intervals from about 30 km. Performance simulations have been based in "middle of the road" transmission figures, and do not include recycling, although the expected gain is a factor of 2.2-2.4. Detailed specs for the aerosol and molecular channel etalons were presented, and transmission budget for each chain, based on experience and projections from GWNH. Design modifications relative to GWNH include the CLIO extender, which uses relay optics to displace the CLIO element from the CCD. This permits CLIO adjustments (alignment) or replacements to be made without risk to the CCD, and reduces thermal load on the CCD to improve background. Custom camera heads will no longer be needed. Balloon winds will have a single camera ‚ will reduce data and need for consumables (such as camera coolant) for balloon mission. Michael identified thermal management as one of the most challenging issues for the flight, since there are large temperature variations at Holloman AFB, and from surface to float altitude. Following a series of reviews, a ground-test version of BalloonWinds is to be ready by 12/04.
Dave Emmitt reported on ground-based TODWL Operations, for which the 2-micron DWL was removed from the aircraft and used to measure cloud and PBL winds from a trailer. A new DWL has been purchased and will be installed permanently on the plane this Fall. In the meantime a "lidarfest" with dual Doppler instruments is being planned. The Series I experiments were conducted on June 9/10 2003 to achieve co-alignment and wind comparison with the VALIDAR instrument at LaRC. Vertical staring and a 7-point, 180 degree VAD section were performed, and anemometers used to validate. Future plans are to compare with HARLIE and other instruments, perhaps a mobile GroundWinds or BalloonWinds system. Cloud and PBL coupling experiments, backscatter and winds correlation, and cloud/chemical lifecycle studies are planned, using the RAPCD facility at Huntsville, AL for the chemical studies.
Michiko Masutani provided an update on NCEP global OSSEs. OSSEs are a scientifically valid approach to guide instrument design, especially cost-capability trade-offs. OSSEs are also valuable in designing an observing system by rigorous testing of the impact of the various components. The NCEP system has been developed with a lot of attention paid to the details of observational data simulation and Nature Run calibration. Preliminary results for a DWL provide an encouraging forecast impact.
Lacey Holland presented results from the Winter Storm Reconnaissance Program (WSRP), with emphasis on adaptive targeting. The WSRP involves multiple government agencies and university collaboration. Targeted observations (dropsondes) are collected over the Pacific based on the anticipation of significant weather events. The program has been operational since Jan 2001, and typically produces a forecast improvement 60% to 80% of the time over downstream regions. Targets are identified, then a sensitivity analysis performed, and finally a fly/no-fly decision is made. When observations are taken, verification is performed for every event. The 54 pre-programmed flight tracks cover a LARGE fraction of the Pacific. While no sensitivity study on selection of flight tracks has been done ‚ since only one can be chosen each time - a null study clearly shows that no improvement is expected by picking a random track over the Pacific. Lacey showed some case studies and statistics. Best results are achieved when adding BOTH winds and temperature. The sondes measure humidity but it is not yet being used. Biggest improvement is found in predictions of surface pressure. Future work includes precipitation verification and threat score studies. The WSRP may be used to support Atlantic and Pacific field campaigns.
Bob Brown presented a paper entitled, "The competition for Satellite Winds: Insider Information." Bob wants the scatterometer experience and information to help DWL community. He noted that PBL rolls that can't be accounted for with models (since they don't have 50 PBL layers) even though organized, non-turbulent roles transporting energy and moisture. However, these large eddies can be detected using scatterometers and fluxes estimated. A list of products other than winds available via the clever analysis of data from scatterometers includes the detection of fronts, storm locations, marine surface pressure fields, surface stress vector, mean PBL temperature, mean PBL stratification, submarine detection, pack ice tracking, and even land surface vegetation. Bob speculates on what DWL might do, based on these lessons: Storm winds are still underestimated. DWL could help improving the climatology record ‚ and predictions. Other targets include the detection of waves and instabilities in troposphere, PBL inversion, the strength and depth of tropopause, and organized vortices. Bob urges that these scientific targets not be forgotten despite the emphasis on numerical weather prediction.
Phil Gatt reported on Recent CTI Advancements in Wind Lidar, with emphasis on 2-micron commercial DWL. CTI has an engineering and an R&D effort, plus a CLR Photonics Division for mature commercial products. CTI has built a ballistic winds program comparable to TODWL. Small systems and containers for their enclosure are available as 4x4 foot packages for remote operation. A movie of PPI and RHI scans made near Dallas was presented showing turbulent eddies and velocity patterns that propagated through the scene. Recent emphasis has been on making these commercial products more reliable for airport customers. Better modular design to allow efficient and fast in field interchange/replacement of components, such as pump diode and MO (which can be field swapped in 15 minutes) has been a key development, as well as improved diagnostics. CTI has also recently installed a side-pumped system on aircraft to support NASA Turbulence Detection and Mitigation activities and the Japanese CDL. So far they have produced about ~ 100 hours of data in test flights, some of which is being used to compile a new 2-micron backscatter database.
Geary Schwemmer presented Tom Wilkinson's talk on Cloud Imagery and Motion. He reviewed the technique of wind determination from MCP (Moving Cloud Patterns) through the block matching technique, with lidar ranging fixing the height and converting angular to horizontal velocity. Improvements made during the last year include "image subtraction" improved cloud motion detection, and measurement by eliminating stationary objects such as persistent sun glare features. Cleaner time series of velocities result. MCP output is highly variable, which results from the absence of visible clouds, or low contrast of cloud features, or having multiple cloud layers visible during an image sequence. Analysis techniques such as time-based filtration and distribution-based filtration are being studied to mitigate this. Contrail analysis is now being automated using the Hough transform approach. In summary, there is a noticeable MCP wind quality improvement due to image subtraction, with efficiency and versatility improved via new software.
Geary Schwemmer (for Charles Earl) went on to discuss HARLIE Winds at IHOP 2002. A group of students working at USU was seeking to calibrate and validate wind lidar measurements, and to extend HARLIE operations to approach 24/7, and support IHOP goals with correlated analysis. For IHOP, HARLIE covered 70% of a 36-day window. This maximized overlap with the other instruments that had lower duty cycles. Altitudes of wind observations varied from 500 m to 15 km. Geary showed good data comparisons (using both automated and manual retrievals), decent ones, and finally not-so-good ones, the last of which appear to be attributable to the automated method perhaps not filtering as successfully and identifying noise as cloud patterns. Even during the "bad" cases, the manual retrieval seems to work, but in the worst case, the automatic algorithm can place the wind 180 degrees in error.
Dave Emmitt (for Dave Bowdle) discussed a "Comparison of Measured and Modeled Aerosol Backscatter at 2 Microns." Backscatter from the TODWL dataset has been examined, and background and convective modes have been modeled to determine whether the prediction for 2 microns based on backscatter at other wavelengths matched observations. The textbook variation predicts backscatter proportional to (Lambda/Lambda_0)**2, but in fact, more like the exponent is ~ 1.4. Using CTI lidar data over land and over sea, greater variability is found over land. Reasonable agreement is noted for measured and modeled maritime. The model seems to be good to first order; captures statistics and vertical structure, etc. Potential refinements include adding relative humidity effects, considering other aerosols, and incorporating the stratosphere.