I am working with monthly-mean MERRA-2 data, specifically the following fields (see p. 39 of the MERRA-2 File Specification document at https://gmao.gsfc.nasa.gov/pubs/docs/Bosilovich785.pdf):
- VFLXCPT, northward flux of atmospheric enthalpy
- VFLXPHI, northward flux of atmospheric potential energy
- VFLXQV, northward flux of atmospheric water vapor
I am using these to compute the monthly-mean, zonally and column-integrated meridional moist static energy flux at each latitude.
I have a question about how these three output fields are produced. Am I correct that (for example) the first one is computed by taking high-resolution (e.g., 6-hourly) data for temperature, multiplying it by high-resolution data for meridional wind, then multiplying by the constant cp (specific heat capacity of dry air), and then integrating vertically over each column on the model's native hybrid coordinates, then finally taking a calendar-month mean?
I am just wanting to make sure that my understanding of how these fields are produced is correct, and that there isn't, for example, any interpolation to standard pressure levels before vertically integrating, or any mass-flux correction step that I'm missing.
Vertically integrated flux fields in MERRA-2
Vertically integrated flux fields in MERRA-2
Hi,
The vertical integration is done at the model's native terrain following coordinate, not the pressure level interpolations. So, cp x Tv(x,y,sigma) x V(x,y,sigma) x DELP(x,y,sigma) , then vertically integrate sigma 0.0 -1.0, the full atmospheric column. The time averaging is done from the model time step and is saved every one hour, then saved at every month (for the monthly mean collection). So, your understanding is correct.
These are computed from the values used in the model predictions, and output directly from the model. They do not have additional mass flux corrections, or corrections for the analysis (as in the work by Mayer or Trenberth).
I hope this helps.
Go to full postThe vertical integration is done at the model's native terrain following coordinate, not the pressure level interpolations. So, cp x Tv(x,y,sigma) x V(x,y,sigma) x DELP(x,y,sigma) , then vertically integrate sigma 0.0 -1.0, the full atmospheric column. The time averaging is done from the model time step and is saved every one hour, then saved at every month (for the monthly mean collection). So, your understanding is correct.
These are computed from the values used in the model predictions, and output directly from the model. They do not have additional mass flux corrections, or corrections for the analysis (as in the work by Mayer or Trenberth).
I hope this helps.
Filters:
-
GES DISC - jimacker
- User Services
- Posts: 21
- Joined: Wed Dec 04, 2024 4:05 pm America/New_York
- Has endorsed: 1 time
- Endorsed: 1 time
Re: Vertically integrated flux fields in MERRA-2
Hello. We are examining your question today. We apologize for the delay.
Re: Vertically integrated flux fields in MERRA-2
Hi,
The vertical integration is done at the model's native terrain following coordinate, not the pressure level interpolations. So, cp x Tv(x,y,sigma) x V(x,y,sigma) x DELP(x,y,sigma) , then vertically integrate sigma 0.0 -1.0, the full atmospheric column. The time averaging is done from the model time step and is saved every one hour, then saved at every month (for the monthly mean collection). So, your understanding is correct.
These are computed from the values used in the model predictions, and output directly from the model. They do not have additional mass flux corrections, or corrections for the analysis (as in the work by Mayer or Trenberth).
I hope this helps.
The vertical integration is done at the model's native terrain following coordinate, not the pressure level interpolations. So, cp x Tv(x,y,sigma) x V(x,y,sigma) x DELP(x,y,sigma) , then vertically integrate sigma 0.0 -1.0, the full atmospheric column. The time averaging is done from the model time step and is saved every one hour, then saved at every month (for the monthly mean collection). So, your understanding is correct.
These are computed from the values used in the model predictions, and output directly from the model. They do not have additional mass flux corrections, or corrections for the analysis (as in the work by Mayer or Trenberth).
I hope this helps.
Re: Vertically integrated flux fields in MERRA-2
Thanks so much, that's exactly what I wanted to know.
As a side note, seeing you as the one answering my question feels kind of like a celebrity sighting, as I've known your name since I first read the MERRA-2 File Specification two years ago. Thanks for your work to make this data accessible!
As a side note, seeing you as the one answering my question feels kind of like a celebrity sighting, as I've known your name since I first read the MERRA-2 File Specification two years ago. Thanks for your work to make this data accessible!