Opened 5 years ago

Last modified 5 years ago

#220 new Defect

Issue with NCEP GFS data after 2015-01-14

Reported by: harish Owned by:
Priority: major Milestone:
Component: FP other Version: FLEXPART 9.0.2
Keywords: Cc: harish.gadhavi@…

Description (last modified by pesei)

We have noticed that the potential emission sensitivity (PES) values calculated using NCEP GFS data (ds083.2) and FLEXPART version 9.0.2 suddenly increases for the date after 14th January 2015. We know that some new variables like 2m dew point temperature are added to NCEP files 15 January 2015 onward but not sure if they are causing the change in model's behaviour. The spatial pattern which we get for PES is fine and comparable between pre and post 14 Jan 2015 except that the values are increased significantly in case of post 14 Jan 2015 runs.

Attachments (4)

pes_2015-01-13.png (323.8 KB) - added by harish 5 years ago.
Potential Emission Sensitivity Map for Ahmedabad 13 January 2015
pes_2015-01-16.png (325.3 KB) - added by harish 5 years ago.
Potential Emission Sensitivity Map for Ahmedabad 16 January 2015
RELEASES (35.3 KB) - added by harish 5 years ago.
RELEASE file
COMMAND (1.6 KB) - added by harish 5 years ago.
COMMAND file of FLEXP

Download all attachments as: .zip

Change History (6)

comment:1 Changed 5 years ago by pesei

  • Description modified (diff)

Thank you for reporting this issue. Would you mind to post some more details about your problem settings?

Which part of the globe? Is there a difference over land and over water? At which source distances is the effect observed, and is the difference a function of particle age or distance?

My first guess would be an effect on ABL parameters such as turbulence and mixing height.

comment:2 Changed 5 years ago by harish

Thanks @pesei
Following are details on the problem.

The new variables introduced in NCEP dataset DS082.3 since 14 Jan 2015 6 hours are following

  • 2m dew point temperature
  • Snow Depth

Following is the list of variables discontinued from 14 Jan 2015 00 hours

  • Mean Sea Level Pressure
  • Total Cloud Cover

Following is the list of variables being read by readwind_gfs.f90 but we don't find mention in the list of variables described on NCEP website

  • lsprec (large scale precipitation)
  • convprec (convective precipitation)

There are few more variables being added to NCEP data from 14 Jan 2015 but we think they may be unlikely to have effect since they are not read in readwind_gfs.f90 file.

  • Apparent temperature
  • Specific heat above 2m
  • CAPE 0 to 250 mbar
  • convective inhibition 0 to 250 mbar
  • Field capacity
  • Haines Index
  • ICAO
  • ozone mixing ratio
  • soil temperature
  • sunshine duration
  • V-wind component 80m altitude
  • U-wind component 80m altitude
  • Ventilation rate
  • Wilting Point
  • Wind speed (gust)

Following is link to NCEP dataset description.
https://rda.ucar.edu/datasets/ds083.2/#metadata/detailed.html?_do=y

We have run the FLEXPART model for a location in India (Ahmedabad city). We attach two plots of PES maps one before 14 Jan 2015 and another after 14 Jan 2015 and COMMAND and RELEASE files of the run. The difference is high near receptor and decreases farther away.

Changed 5 years ago by harish

Potential Emission Sensitivity Map for Ahmedabad 13 January 2015

Changed 5 years ago by harish

Potential Emission Sensitivity Map for Ahmedabad 16 January 2015

Changed 5 years ago by harish

RELEASE file

Changed 5 years ago by harish

COMMAND file of FLEXP

Note: See TracTickets for help on using tickets.
hosted by ZAMG