Docstoc

11-29pm_Leblanc_IntroTalk

Document Sample
11-29pm_Leblanc_IntroTalk Powered By Docstoc
					Welcome




      2010 ISSI International Team of Experts on:

          Critical Assessment and Standardization
                      of the Definitions
          of Vertical Resolution, and Uncertainties
  in the NDACC Temperature and Ozone Lidar Algorithms

          (and its reporting in the NDACC data files)

                               Thierry Leblanc, JPL




             2010 ISSI Team on Lidar Algorithms, First Meeting, 29 Nov. - 3 Dec. 2010, Bern, Switzerland
                       Why bother with vertical resolution?
NDACC
Context




    1. Time (therefore vertical) discretization by lidar is not infinite
    For NDACC systems, typically runs between 15-m and 600-m



    2. Lidar Signals do contain noise
    That is why we do smooth/filter



    3. NDACC data users often interpret observed vertical features
    Examples: Gravity waves, ozone filamentation, stratospheric warmings



                Yes, we do need to bother !




              2010 ISSI Team on Lidar Algorithms, First Meeting, 29 Nov. - 3 Dec. 2010, Bern, Switzerland
                               Why bother with uncertainties?
NDACC
Context



      1. Lidar signals do contain experimental noise
      Typically, photon-counting noise following Poisson statistics



      2. Lidar signals are subject to systematic errors
      Typically, due to non-linearity, overlap, a priori (e.g., cross-sections), etc.




      3. NDACC data users need information on accuracy
      Example: why is there a 5% difference between
      Mr. Smith’s lidar and Mr. Jones’ ozonesonde measurements?


                 Yes, we do need to bother !




               2010 ISSI Team on Lidar Algorithms, First Meeting, 29 Nov. - 3 Dec. 2010, Bern, Switzerland
                                   2010 ISSI Team = Hard work!
    Why                                      (sorry)
are we here?



        What happens usually at small-attendance meetings and workshops?

           1. A lot of talking
           2. A little bit of (or no) acting
           3. A lot of “action item” setting



        What should we expect from our 2010 ISSI Team?

           1. A little bit of talking
           2. A lot of acting
           3. A little bit of “action item” setting



                       Are you ready for this?



                 2010 ISSI Team on Lidar Algorithms, First Meeting, 29 Nov. - 3 Dec. 2010, Bern, Switzerland
                              To have all NDACC lidar PIs
 What is                   report data/metadata consistently
our goal?




      1. ALL NDACC PIs eventually should report vertical resolution
         based on the same definition(s)




      2. ALL NDACC PIs eventually should report uncertainties
         based on the same definition(s),
         and on the same rules of error propagation




              2010 ISSI Team on Lidar Algorithms, First Meeting, 29 Nov. - 3 Dec. 2010, Bern, Switzerland
                             by using as practical of an approach
How can we                               as possible
achieve this?



       1. Review/assess key points on the topics
       Not a thorough review, but instead, a review adapted to our needs



       2. Develop tools to make the necessary changes practical to PIs
       Easy-to-use “packages”, in multiple languages, to accommodate all



       3. Detailed and accessible documentation
       Documentation must be easy-to-use, easy-to-access



       4. Enforce implementation
       Sustained interaction with all PIs to insure successful implementation




                  2010 ISSI Team on Lidar Algorithms, First Meeting, 29 Nov. - 3 Dec. 2010, Bern, Switzerland
                                     Remain focused on our task!
What shall we
do this week?


        1. Provide detailed description of our own filtering procedures
           and uncertainty estimates
           What filter? What window? How many times? Where in software?
           Error sources: How many? Where? How much? How do they propagate?


        2. Synthesizing, then define the best approach towards standardization
           Think collectively, not individually; Think as a data user


        3. Split in sub-groups (if necessary) to increase focus and efficiency
           Sub-group on smoothing filters? Sub-group on derivative filters?
           Sub-group on error source estimate? Sub-Group on error propagation?


        4. Contribute actively to tool implementation
           Pick your programming language(s), and develop and/or implement tools


        5. Wrap up and prepare for next steps (homework, meeting 2)
           Define clear timeline, and action items, for after meeting 1


                 2010 ISSI Team on Lidar Algorithms, First Meeting, 29 Nov. - 3 Dec. 2010, Bern, Switzerland
The end




                           No more talking,
                           Let’s do it…




          2010 ISSI Team on Lidar Algorithms, First Meeting, 29 Nov. - 3 Dec. 2010, Bern, Switzerland

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:11
posted:8/4/2012
language:English
pages:8