51100849a96f26570155809c20a509f9.ppt
- Количество слайдов: 12
Automated Operational Validation of Meteorological Observations in the Netherlands Wiel Wauben, KNMI, The Netherlands l l Introduction QA/QC chain Measurement system and users Status
Introduction l l Automated network for synop and climatological observations. Data near real-time available to internal and external users every 10 -minutes. Observers at airports only for aeronautical reports, but 12 second wind and RVR data provided continuously. Automated network requires automated validation in real-time.
QC chain 6 months, technical & station Calibration period 8 -24 months or problems, allowed range for deviation Range, jump persistency, basic interrelation Instrument selection Procedures Interrelations, temporal, spatial Off-line, daily Reporting vs sensor errors, Handling of quality information Real-time?
Data flow (Met. Net)
Basic assumptions l l l l 24*7 considered usefull and reduces manual labour “No” delay in data flow QC does not change values Result of QC check in binary Q-flag Manual input (link to technical/environmental changes) Alarm Validation results should be embedded in QA/QC chain with suitable actions to eliminate causes
Follow up l l l l Overview current QC at various places Details of methodes and usefullness (number, importance) Optimal location of QC (OMWA, 10 min) Q indicators traceable throughout data flow (sensor-interface BUFR report) Follow up (e. g. single jump in temperature) User should use data AND quality (mask applied for the users Start with Met. Net but keep general
Ceilometer (NI, QG and statistics)
Ceilometer statistics
Radar versus precipitation gauges l l l Scatter plot Daily sums Dependent verification since bias is removed
VIMOLA vert. integr. LAM l l l Quasi geostrofic P at msl 10 m wind currently short term forcast using hourly data “any” resolution indicates suspect P values
Current valiation (daily, non-RT)
Outlook l l Make business case for basic 10 -min near real-time validation Investigate other possibilities for temporal, spatial and interrelations in RTV l QC at other NMI’s l Start implementation of basic version l Allow for extensions/generalisation
51100849a96f26570155809c20a509f9.ppt