Top

GPS-Met System Monitoring      

April 24, 2018

Information and Links

Current Status Time, #Stations Interactive Display TPW maps
Interpolation
TPW maps
NOAA color

Update MM-DD, HH:MM


Current # of stations

> Detailed Status

Verification Comparison
RAOB and SUOMINET
O-B
GFS 12hr fct
System Info. Flow & Data



Planned Updates and Maintenaces

DateTime Actions
None


System Problems and Updates (which affect product quality and delivery)

Problem Data Quality   Data Missing > 6 hours <= 6 hours     Current Problem   Updates Improvement

Date Problems and Actions Fixed / Links
March 26, 2018
( Potential Problem )
No METARS Data : Missing periods for TPW products are expected because of missing METARS data.
March 13, 2018

Web site for GPS Solutions have been updated.
October 7-10, 2017
( Potential Problem )
Scheduled Maintenance in NOAA Servers : Some missing periods are expected in case of the problems in delivery of METARS data.
October 5, 2017

Missing of GPS-MET products for three-four hours: No GNSS products were not availble from around 06UTC for 3-4 hours because of upgrade of hosting service. The problem was fixed by setting optional hosting service.
October 3, 2017

Missing of GPS-MET products for one-two hours because of no METARS data: No GNSS products were not availble because of no surafce meteorological observation data from NOAA AWC (All of metars csv files are empty only with header in AWC) METARS data has been abailble since around 19:20UTC, and GPS-MET system has delivered the products.
Septeber 2-5, 2017 Some Raw Data Missing: No GNSS data were available from NOAA NGS CORS network (about 700 - 800 stations). GPS-MET system tries to download GNSS data for four times, but no data was available for about 3 days over Labor Day weekend. GPS-MET system was woring with data from other data providers (UNAVCO, SOPAC, TX DOT, ...) in the periods. Raw data in FTP site in NOAA NGS CORS has been available from 14UTC, Sep. 5
July, 2017 General: Statistics of GPS-TPW minus RAOB-TPW is updated (July, 2017).

Link
July 13, 2017 Product missing for a few hours: Products were not available for the period 11-15 UTC, July 13, 2017 because of temporal system problem in data server. The problem has been fixed after system oprator got error message from the system.
July 1-5, 2017 Some Raw Data Missing: No GNSS data were available from NOAA NGS CORS network (about 700 - 800 stations). GPS-MET system tries to download GNSS data for four times, but no data was available for about 4.5 days. GPS-MET system was woring with data from other data providers (UNAVCO, SOPAC, TX DOT, ...) in the periods. Report
June, 2017 General: Statistics of GPS-TPW minus RAOB-TPW is updated (June, 2017). Eleven USCG stations started sending GPS data from late in June.

Link
May, 2017 General: Statistics of GPS-TPW minus RAOB-TPW is updated (May, 2017).

Link
April 18-21,
2017
Problem: GPS-TPW showed dry bias

Reason: It was caused because GPS-MET system used one incomplete Veripos satellite products (because of system update) with missing of orbit and clocks (only header). The monitoring program prepared on the April 5 action thought the files existed there, and no alert were sent. New GPS-MET processing system (RTNet) were available at that time, but it was still used for comparison purpose (parallel processing). That caused generation of biased GPS TPW products, again, during the second problematic period (April 18-21).

Updates:
(1) New program which pickup optimal Veripos satellite products were made, and it now choose complete products for GPS-MET processing.
(2) One more operational Veripos system as additional source of GPS orbit/clock products.
(3) RTNet software was switched to the latest one (which had been tested in parallel processing) on April 21. The new RTnet uses better approach than that in the action (3) (Add more statistical noise (sigma) to the ZTD estimate). The new RTNet doesn't output any ZTD solutions when GPS orbit/clock products are not available.
With these updates, TPW products will not be generated and sent in the case of no Veripos satellite products problems on April 5 and April 18-21.
April 21,
2017
April 5,
2017
Problem: GPS-TPW showed dry bias

Reason: Someone removed the file needed to authorize copying via "rsync" in primary Veripos data server. GPS-MET system continued to output ZTD estimate for the last epoch before missing of Veripos satellite products. This is the first event of missing of Veripos products since October 27, 2016.

Updates:
(1) Two more independent system are used to download Veripos satellite products.
(2) Send notification in case of missing Veripos satellite products to monitor that the redundant systems are working correctly.
(3) Add more statistical noise (sigma) to the ZTD estimate in our internal ZTD output during periods when GPS orbit/clock are missing. Should the noise in the ZTD estimate exceed a set threshold the GPS-MET QC system automatically rejects biased ZTD data based on the noise level.
(4) TPW data are compared to NCEP GFS internally in addition to radiosonde TPW.
April 5,
2017
October 27,
2016
Update: Change of satellite products from regional to Veripos global

Reason: Because ZTD estimate with regional satellite clock products are noisy in the stations (e.g. Hawaii) far from the server network used for satellite clock estimation, satellite clock are updated to Veripos global products updated every 60 minutes.

Improvement:
(1) ZTD and TPW products are improved (better agreement with RAOB TPW in Hawaii stations).
(2) No more dependency on unstable FTP data download from free GPS data.
--


------------------------------------------------------------------------------------------------------------------------