Dear CAMS user support (cc @Michela),
We use the concentrations of the dataset CAMS global inversion-optimised greenhouse gas fluxes and concentrations (cams-global-greenhouse-gas-inversion, ADS) as initial and boundary conditions for our forward simulations using ICON-ART. For that purpose, we need the geometric height of the model levels at each time step. That task is quite easy for the CH4 product (TNO), given that hybrid levels, surface pressure, temperature and specific humidity are within the dataset. Moreover, for CH4 the time-dependent geometric height field is also given as part of the dataset. However, for CO2 and N2O we have some problems.
What we observed, e.g., for instantaneous concentrations, surface-driven inversions, January 2023, region 48–52° N, 8–12° E:
- N2O v23r1 and CO2 v23r1 and v25r1 have the same 79 layers: their ap and bp are identical.
- Their horizontal grids differ. N2O is on a regular grid of 0.703125° × 1.40625°. CO2 is on a 1° grid, “interpolated from the hexagon grid of the transport model (~90 km)”.
- The CO2 files contain height_above_reference_ellipsoid at the 80 layer interfaces, with no time dimension. The N2O files contain no height.
- Neither product contains temperature or humidity. The interface pressures can be computed from ap, bp and Psurf, but the geometric height cannot, given that geopotential cannot be calculated.
- Older versions (CO2 v17r1, N2O v16r1) have 39 layers, and their Psurf is described as “Dry air surface pressure”. In the recent versions it is described as “Surface pressure”.
Questions for LSCE (CO2):
- What does the time-independent height_above_reference_ellipsoid represent now? For instance, is it a temporal mean, a climatology, or the height for a given date? With which surface pressure and temperature was it computed? Is it usable for the vertical interpolation?
- Which meteorological fields drive the transport model (e.g., ERA5), and which variables are nudged (winds only, or also temperature)?
- Would it be possible to provide a time-dependent geometric height of the interfaces or cell centers, or the temperature and specific humidity on the model levels?
- Is it reasonable to use this height for other products on the same 79 levels, such as the N2O product, after horizontal interpolation?
Questions for NILU (N2O):
- Which meteorological fields drive the transport model, and which variables are nudged?
- Could future versions include a time-dependent geometric height of the layer interfaces or cell centers, or the temperature and specific humidity (as the CH4 product does) so that one can calculate the geometric height accordingly?
Questions for both teams:
- Does the time stamp of the instantaneous fields mark the start, the end or the centre of the 3-hour interval, or an instantaneous state? Why do the two products differ by 3 hours?
- Is the change of Psurf from “Dry air surface pressure” to “Surface pressure” a change of content, or only of description? In which versions did each product change from 39 to 79 layers?
Thank you very much for your help.
Kind regards,
Dr. Diego Jiménez de la Cuesta Otero, DWD