0a5d03c9250ba4c6d9cb014d9e1cb471.ppt
- Количество слайдов: 23
RDA Survey 2005 CISL Executive Meeting 1 August 2006
Outline • • • Basics General Quantitative Summary Response to Three Questions in more Detail Who is using what from the RDA? Assessment of Open-ended Questions – Mapping to Work Underway and Futures • How to gain more from user surveys
Basics • User pool – Registrants on RDA online server • No restrictions on who can register • Must register to get data - general information (metadata) does not require registration. • Year group 2005 only, 3914 invited to survey • Respondents – 476, about 12% – Open period, 45 days (4/27 6/12)
Basics • Divide user community by affiliation – Enables response evaluation for separate groups Observations and Assessments • ~ 60% are Non-U. S users + international hub for data 180 + supports policy of ‘free and open access’ + leverage in forming international data exchanges - use resources (hardware and network) to serve world 98 71 38 30 22
Basics • Seven questions with fixed responses – Choose one of five levels of satisfaction • Very Dissatisfied, Neutral, Satisfied, Very Satisfied • Simple assessment, add (N + S + VS), % satisfied 97%, Content of the RDA for your research (410) 95%, Accessing RDA data from the DSS server using FTP or web browsers (405) 90%, Accessing RDA data directly from the NCAR Mass Storage System * (215) 96%, RDA metadata, documentation, and supporting information from the DSS server (369) 94%, Data specifically prepared for you by the DSS staff, including FTP downloads, tapes, and other media (227) 95%, Consultation and assistance from the DSS staff (224) 96%, Discovering RDA data using search, catalogues, and project information from the DSS server (353) (nnn) = total respondents * Potentially some bias?
Observations and assessment • single strong mode - S • would like to see stronger VS, especially for US University
Observations and assessment • good strength in S and VS, users like simple access • not much variability between affiliations, good US University approval
Observations and assessment • good strength, and not much difference between S and VS • high variability between affiliations in VS, good US University approval
Who is using what? Observations and assessment • generally users tap the RDA for multiple types of data • good balance in the major categories, led by reanalyses
Assessment of Open-ended Questions • Three questions - generated ~ 50 pages of text • Many positive comments, good, - affirms our efforts • But, we want more! – assess the comments that had suggestions, collectively, map to DSS developments underway and determine relevant future directions • Process – Four staff and SJW read, digested, and discussed - look for the bigger message - very subjective – Didn’t over analyze - many interpretation are possible
Assessment of Open-ended Questions Question 1 • List your preferred methods for discovering data (e. g. Google search, local site search, catalogs, etc). – Weak, leading question – All modes are preferred, plus, directives from colleagues, and discovery from other websites • Observations and actions – Users can now use the UCAR Google search from the RDA website, since March 2005 • Restricted to RDA results • Need to strengthen ‘meta tags’ on RDA documents – Some are not finding data that we do have (comments and small VS on bar chart) – Well into a metadata project with features - 18 months • Tightly constrained standard - GCMD with mapping to CF • Database (My. SQL) stored and accessible • Enables fast and accurate browse and search • Allows progressive deployment on a large archive • Roll out in 2006
Assessment of Open-ended Questions Question 2 • List any datasets your projects need that should be added to the RDA. – Higher spatial and temporal resolution - everything! – More near real-time – Atmospheric • NEXRAD precipitation mosaics for CONUS • High resolution global forecast winds • Chemical components, aerosols • Time series indices (AO, NAO, etc) – Ocean • Reanalyses • Weekly SST, monthly SSS • Ocean bottom composition • Sea ice (thickness, velocity) – Satellite • TMI, TRMM (precipitation) • Jason-ERS-TOPEX • GOES Images – Hydrology • River flow • Soil moisture • Land surface, vegetation
Assessment of Open-ended Questions Actions • Added NCEP Global Ocean Data Assimilation System Reanalysis, 1980 -> • Will added higher resolution atmospheric datasets (analyses and forecasts) – TIGGE – Improvements to IDD derived archive (real-time data delivery) Current (161 MB/day, 50 GB online) Model RUC NAM GFS ENS ECMWF Coverage CONUS N Hem. Global N. Amer. Global Resolution Init time 81 x 81 km Hourly 81 x 81 km 00, 12 381 x 381 km 00, 06, 12, 18 5 x 5 deg 00 1. 5 x 1. 5 deg 00/12 5 x 5 deg 00/12 New (17 GB/day, 500 GB online) Model Coverage Resolution Init time RUC CONUS 40 x 40 km Hourly RUC 2 CONUS 20 x 20 km Hourly NAM CONUS 12 x 12 km 00, 06, 12, 18 GFS Global 0. 5 x 0. 5 deg 00, 06, 12, 18 GFS Global 1. 0 x 1. 0 deg 00, 06, 12, 18 GFS Global 2. 5 x 2. 5 deg 00, 06, 12, 18 ECMWF Global 5 x 5 deg 00/12 Length/Interval 12 h/3 h 60 h/12 h 120 h/6 h 240 h/12 h 84 h/6 h 168 h/24 h Size 53 mb/day 23 mb/day 11 mb/day. 19 mb/day 69 mb/day 5. 2 mb/day Length/Interval 12 h/1 h 12 h/3 h 84 h/3 h 180 h/3 h 192 h-384 h/12 h 168 h/24 h Size 490 mb/day 3. 4 GB/day 2. 0 GB/day 4. 4 GB/day 6. 5 GB/day 300 mb/day 5. 2 mb/day
Assessment of Open-ended Questions Other Actions Discovery: From improved metadata, create ‘better’ topic specific catalogs • Example, complete prioritize reference catalog for RDA precipitation datasets • Related references at other sites, e. g. TRMM Satellite data at NASA • Potentially, share catalogs via OAI-PMH (Open Archives Initiative Protocol for Metadata Harvesting) Generally, put more data online • FNL Analysis (very popular with MMM and WRF users) – Currently one-year, boost to full record - starting in 1999 • Others - when action supports many users and helps DSS Stay proactive in building content • Status quo (search out additions, collaborate on data projects, etc) • Open an online user suggestion box?
Assessment of Open-ended Questions Question 3 • Is your science limited by the research data available? If yes, what data or services are needed? • Issues that confuse this user group – Data access methods • RDA server, MSS, special requests, SCD data access computing accounts, CDP – Cost for data • Much free, some not, various legacy policies, differences across affiliation – Site navigation Solution (in part), reduce fragmentation of the information across the web site, create clear concise documents in a single place, bring navigation to modern look and standard.
Assessment of Open-ended Questions Question 3, continued • Is your science limited by the research data available? If yes, what data or services are needed? Wish List • All data available in ASCII and net. CDF • More direct linkages with field program data - need better UCAR-wide discovery (CDP). • Easy access to older datasets, which are generally not online • Direct access to the MSS • Formats that can directly couple to regional and global models
How to gain more from user surveys • Transition to surveys designed by professionals – RDA questions are too ad hoc – There is much speculation in the interpretation • The discussions are interesting but summary is very subjective – Want metrics for two subjects • Past usage satisfaction • Quantifiable clues for the future • Need to apply correct statistical tests – Account for sample size – Determine confidence intervals and/or thresholds
Summary • Very pleasing overall results for the RDA • Our developments are aligned to address the bigger issues, in most cases • Other suggestions give us awareness for future work • The RDA is a node of international data collaboration • Certainly will not run short of tasks to work on in my life time.
Some people like us! Underlines by SJW • I have been extremely happy with the data and services provided. Your organization is an efficient, friendly, affordable and responsive alterative to NCDC. I have absolutely no complaints. Keep up the good work • Very good and friendly service. I consider year back global free data as a big achievement especially if to compare with ECMWF or with NZ meteorological data services. It's very helpful for small research groups all over the world. Good luck • Great selection of browsable datasets. Have used this over the past few years to locate hard-to-find data elsewhere • I am very happy with the service and also the response from the staff. At one occasion…. . . I got a very prompt and helpful response. I hope such an excellent service can continue to be supported • A very good free service that we in Europe could not do without. I only wish that our own institution provided the same service • The US philosophy of open access is very commendable. Good work! • Absolutely no complaints a great service and a great benefit to the wind industry. well done! • Very good work. I am satisfied with the service and the data and consider that the services are highly needed by the scientific community. Thanks