Upload
others
View
3
Download
0
Embed Size (px)
Citation preview
Principal Investigator:Judd Welton, NASA GSFC Code 612
Network Manager & Engineer:Sebastian Stewart, SSAI GSFC Code 612
Data Processing:Data Manager: Vacant
Larry Belcher, SSAI GSFC Code 612
Science Team:James Campbell, Naval Research LabJasper Lewis, UMBC GSFC Code 612
Simone Lolli, CNR, Italy
Administrative Support:Lisa Nalborczyk, SSAI GSFC Code 612
CALIPSO Validation Activities:Judd Welton, James Campbell
AERONET & Synergy Tool Partnership:Brent Holben, NASA GSFC Code 618
Dave Giles, NASA GSFC Code 618
NASA SMARTLABS Field Deployments:Si-Chee Tsay, NASA GSFC Code 613
Site Operations & Science Investigations…. many network partners around the world
MPLNET is funded by the NASA Radiation SciencesProgram and the Earth Observing System
The Micro Pulse Lidar Network (MPLNET): 2000 - current
Windpoort, Namibia (borders Etosha Nat Park)Photo: Seb Stewart
Ragged Point, BarbadosPhoto: Seb Stewart
The Micro Pulse Lidar Network (MPLNET): 2000 - current
• Objective: Long term, global lidar network to profile aerosol and cloud vertical distribution and properties
• Provide both NRT and Quality Assured data in support of various NASA missions, programs, and modeling • Support other domestic and international aerosol & cloud research activities• Develop capability to support aerosol modeling (Aerocom) and operational forecasting (ICAP, WMO SDS-WAS)
• MPLNET is a federated network built on the AERONET model: utilize standard instruments, calibrations, operations, and processing• MPLNET is a member of the WMO Global Atmospheric Watch (GAW) Aerosol Lidar Observation Network (GALION): Welton co-chair
MPLNET Stats: 2000 – current• Staff: 612: Welton, Stewart and Belcher (SSAI), Lewis (JCET). Offsite: Campbell (NRL), Lolli (CNR), and numerous site PIs worldwide• 50+ sites, some no longer active• Currently: 20 long term sites (19 upgraded with polarized lidars)• Nearing completion of our new Version 3 system with polarized MPL, and enhanced data products.
• NRT data available on website (<1.5 hrs) including some with QA. Interactive data browsing.• NETCDF 4, CF compliant formats. Subsets available (including custom as requested, Aerocom?).• Automated, hourly FTP product staging for ICAP/SDS-WAS applications
• http://mplnet.gsfc.nasa.gov
SPARTAN and other collaborations:
MPLNET is moving into a 3rd phase, or Version of operation (similar to AERONET)
Developing a more mature & robust instrument and data suite• Capable of more operational deployments managed under contract to commercial entity• This requires MPLNET to improve and standardize our protocols, calibrations, data processing
This will benefit and enhance collaborations with other networks, ie SPARTAN
Active
Planned
Planned(former site)
Inactive
Active
Planned
Planned(former site)
Inactive
Proposed* revive Ragged Pointand Cape Verde* add Suriname
The Micro Pulse Lidar Network (MPLNET): Sites 2000 - current
Tropics & Mid-Latitudes
High Lat & Arctic
Antarctic
MPLNET - SPARTAN Colocations:
Kanpur
Singapore
Hanoi (former site)
Manila (new, June install)
Pretoria (proposed, SEALS-sA partner)
Bandung (close to planned site, 7SEAS partner)
Updates: V3 remains in beta mode, not fully operational yet
• V3 Product Development• L1 Signal, Cloud, and Aerosol products are complete• L1 PBL product is moving to beta testing• L15 and L2 QA parameters and protocols
• Cloud product is done• Aerosol product nearly complete• Signal product QA was done at last meeting, revisiting this issue!
• L15 and L2 product development will proceed quickly once QA process is done• New workstation purchased, more CPU for multi-threaded processing
• V3 Instruments• All but one site in network have polarized MPL now• Polarization calibration protocol complete (this was a big effort ongoing at last meeting)• New QA issues arose upon inspection, now working on characterizing other components
(main delay with V3 progress)• Testing and evaluating the miniMPL instrument for support in MPLNET
• Personnel• Lead programmer left in summer 2006, replacement hired (start date ~June)• Plan to shift network management to MPL manufacturer under new contract
• Too difficult to continue this part of the project onsite at GSFC
MPLNET: Updates since last meeting I attended in 2016
…. We also have a new president
Signals
Vol Depol Ratio
Laser Energy (u
J)TemperaturesLaser Energy
Solar Background
MPLNET V3 Data Samples: Level 1 Signal Product (L1_NRB)
Lidar Change
Vol Depol Ratio
Signals
Cloud MaskCloud Phase (new) WaterIceMixedUnknown
Thin Cloud Optical Depth (new)
MPLNET V3 Data Samples: Level 1 Cloud Product (L1_CLD)
Aerosol Backscatter
Aerosol and Thin Cloud Extinction
Aerosol Depol Ratio (new)
AERONET AOD MPLNET AODThin Cloud Optical Depth
MPLNET V3 Data Samples: Level 1 Aerosol Product (L1_AER)
Volume Depolarization Ratio
Signals
MPLNET Version 3 Products: L1 PBL (beta testing)
Lewis et al., Improved boundary layer depth retrievals from MPLNET, JGR, 2013
Determines up to 3 potential PBL HeightsAlgorithm selects “most likely” PBL Height, others provided in product
Lewis co-I on NASA ACCDAM project to improve PBL parameterization in WRFResults will help refine the operational implementation of the PBL research algorithm
5 minute signal average
Polarized Micro Pulse Lidar
MPLNET Version 3 MPL Characterization: polarization and beyond …
• The polarization characterization was completed by early 2017. New calibration protocol and QA related parameters have been developed.
• Once the new V3 aerosol products were in heavy beta testing we noticed issues in data, namely day/night bias in signals.• Primarily affects our continuous aerosol products• AERONET constrained retrievals to a lesser extent• This lead to a hold on all V3 progress about 1 year ago as we began
working the problem.
• The problem is two-fold, all driven by temperature changes.
• Changes in the lidar overlap change the profile shape in lower troposphere• Changes in receiver efficiency alter the signal magnitude
• These problems have been known for years but new MPLs seem to be more sensitive
• We have been characterizing the temperature dependence of nearly all optical components: Telescope, Window, Filters+Laser, Detector
• We are still finalizing this work. Will lead to changes in our L15 and L2 QA process. No change to L1.
Uniform Heating/Cooling of Telescope
-10 DC
+10 DC
Telescope cannot accurately image near range signals to detector. Resulting signals too low, large correction required!
Signals past full telescope overlap range are accurately imaged to detector, no correction required.
Window Radial Temperature Gradient: center to edge
Worst Case: Uniform Telescope Heating/Cooling , plus …. Non-uniform Scope Temp Heating/Cooling and/orWindow Radial Temperature Gradient and/orOpto-Mechanical Alignment Imperfections
New Overlap Simulation Capability:
Expanded on original method from Berkoff et al 2005• Only addressed uniform telescope heating/cooling• Not formulated for operational processing (R&D)
Improvements:• Fixed some bugs• Added telescope obscuration from secondary mirror• Added Window temperature gradients (axial & radial)• Added opto-mechanical alignment imperfections
• ex: pinhole offset from ideal location• Operational implementation being developed
MPLNET Version 3 MPL Issues: Telescope & Window TemperatureOverlap changes alter signal shape (aerosol retreivals most affected)
Enclosure Air Temp
Detector Temp
Laser Temp
Solar Background
Signal at 9 km
MPLNET Version 3 MPL Issues: Temperature and Receiver EfficiencySignal Magnitude Problems
This type of day is especially difficult with an offset between solar noon and afternoon heating.Telescope, window and laser+filter issues depend only on temperature, but are 2nd order effects for signal magnitude problems.The photon counting detector behavior is more complicated, and we think a primary cause of the issues above.
AERONET AOD
MPL AOD
MPL Calibration Value(count rate to SI unit)
MPLNET Version 3 Instrumentation: testing and evaluating miniMPL
• The MPL was developed at NASA GSFC and patented in early 1990s
• The current manufacturer of the commercial MPL is Sigma Space (now Leica), in place since ~2005.
• The miniMPL was developed by the current MPL manufacturer several years ago as an offshoot of the MPL, without NASA assistance.• The miniMPL is a different optical design and current MPLNET calibration protocols cannot be applied.• The instrument has not been rigorously evaluated and characterized.• Despite this issue, many miniMPLs have been sold worldwide, probably more than MPL now.
• Support of this product in MPLNET has been on hold till now.• We are evaluating its performance and will develop calibration and QA protocols suitable for its use.• I purchased one, running at GSFC since last year.• Level 1 signal data (NRB) are suitable for MPLNET and approved now. Higher level products TBD.• Several new partners with miniMPL will be added to network over coming year
GSFC MPL (operational)
MPLNET Version 3 miniMPL testing: product retrievals
Last Week at GSFCGSFC miniMPL
Lower miniMPL signal-noise ratio: Daytime depolarization data severely degraded above PBLV3 Cloud detection fails to detect thin cloudsV3 Aerosol algorithm fails to detect weak aerosol layers
Conclusion:
We continue to plod on …
• All V3 L1 processing code should be finalized by summer.
• We have already begun recalibration of all data to L1 V3 standards
• Reprocessing of all data to V3 will begin soon, dependent on getting our new workstation setup• Current processing computer lacks CPUs needed for network size• New programmer & sys admin to start soon will help setup• already testing new workstation• Cuts hourly L1 processing from ~20-30 mins to < 5 mins, with CPU to spare for other work
• Network continues to grow, with at least 3 new sites to come online by end of 2018
• GALION (GAW Aerosol Lidar Observation Network)• Work will begin this summer on a GALION data center
• Website front at GAW utilizing distributed data center infrastructure (US & EU)• Lidar network data and site discovery tools• Eventually data imaging and serving• Calibration and QA support using agreed upon community standards
• Discussions with UMBC and Howard U colleagues to create a GALION calibration facility here in Maryland (to serve the US, perhaps other Western Hemisphere regions)• Would be in cooperation with ongoing European calibration efforts