Automation Schedule (VPC) (deprecated)


A virtual PC (VPC) running Windows 10 is used for the daily automatic calculation and generation of ecosystem fluxes and QuickViews. Below is the schedule that shows at which time fluxes and QuickViews for the respective site are generated.

All times are given as CET. The VPC is running on winter time.

wdt_ID Site GA Housekeeping Flux Calcs Merging QVs
1 CH-OE2 IRGA72 01:30 05:10 07:30 08:40
2 CH-LAE IRGA75 01:25 05:00 07:30 08:35
3 CH-FRU IRGA75 01:15 05:40 07:30 08:30
4 CH-AWS IRGA72 01:00 05:20 07:30 08:15
5 CH-CHA IRGA75 01:05 05:30 07:30 08:20
6 CH-DAV IRGA72 (ETH) 01:10 05:50 07:30 08:25
7 Dashboard all 08:45
8 CH-INO IRGA75 01:20 04:50 07:30 08:10
9 CH-CHA QCL 01:05 06:10 07:30 08:20
10 CH-INO LGR 01:20 06:20 07:30 08:10

Description:

  1. GA … gas analyzer for which the automatic eddy covariance fluxes are calculated.
  2. Flux Calcs … time at which the automatic calculation of the newest eddy covariance fluxes is started.
  3. Merging … time at which the automatic flux results for all sites from 2 are merged into one file. An additional flux file for each site is created that contains the newest flux results, ready to be used in the QuickView generation.
  4. QVs …start of QuickView generation.
  5. Housekeeping … previous flux results are deleted to make room for the new flux results.

 

What is happening here? Let me give you an example for CH-AWS

During the night, eddy covariance (EC) raw data arrive at ETH and are stored on the grasslandserver (4:30). Soon thereafter (5:20), the flux calculation script for CH-AWS [1] is started via the Windows Task Scheduler. It searches for recent EC raw data on the grasslandserver and uses it to calculate fluxes for the last days. Since fluxes for only a few days are calculated, flux calculations are fast and take only approx. 30mins, which means that the flux results for CH-AWS are ready by approx. 5:50. A subset of the flux results (based on the EddyPro *full_output* file) is then created and added to previous flux results that are stored on the group drive [2]. This means, file [2] is collecting results from all flux runs in one file, i.e. this file is growing longer and longer. This results subset for CH-AWS is then merged with the subsets from all other sites into one file [3] at 7:30.

The script for the generation of the QuickViews for CH-AWS is starts at 8:15. It searches for newest flux results [2] on the group drive and also for newest raw data files (including webcam pictures) on the grasslandserver and uses these data to generate the plots for CH-AWS [4]. At 8:45, the CH-AWS data are also used to generate the Dashboard on the QuickViews homepage [5].

Finally, the following night the most recent flux results are deleted from the group drive. Note that file [2] is not deleted but kept since it is needed to collect all recent flux results in one growing file.

References

[1] P:\Flux\VPC_automatic_fluxes\vpc_automatic_flux_calcs\CH-AWS\CO2_H2O_IRGA72\start_FCT.py
[2] P:\Flux\VPC_automatic_fluxes\vpc_automatic_flux_calcs\CH-AWS_current_fluxes.csv
[3] P:\Flux\VPC_automatic_fluxes\vpc_automatic_flux_calcs\CH-SFN_current_fluxes.csv
[4] http://www.swissfluxnet.ch/quickviews/alp_weissenstein/ (now deprecated)
[5] http://www.swissfluxnet.ch/quickviews/swiss_fluxnet/ (now deprecated)

Additional Information

Automatic fluxes are calculated for the past 2 days. Since the eddy covariance data files are transferred to the ETH grasslandserver during the night @04:30h, first calcualtions start soon thereafter @04:50h. Since fluxes are calculated for only few days, calculations are fast and take approx. 30mins per run and site.

Since the upgrade of the VPC (see here) it is possible to run flux calcs on a tighter schedule

 

Last Updated on 16 Feb 2024 00:02