GEOS57_nv_np

GEOS57 provides the data at both native grid inst3_3d_asm_Nv at 72 hybrid levels and pressure grid inst3_3d_asm_Np at 42 pressure level.

Pressure as a function of geopotential height from Np and Nv data

Both datasets have surface pressure and geopotential heights. Since both datasets are derived from the same model, one can expect that the atmospheric pressure as a function of geopotential height will be the same within rounding errors. Unfortunately, this is not the case.

I took DAS.ops.asm.inst3_3d_asm_Nv.GEOS572.20110920_0000.V01.nc4 and DAS.ops.asm.inst3_3d_asm_Np.GEOS572.20110920_0000.V01.nc4 files. I extracted DELP(pressure_thickness), P(mid_level_pressure), H(mid_layer_heights), and PS(surface_pressure) fields from the Nv data product. I extracted H(edge_heights), PS(surface_pressure) and lev from the Np data product.

PS[Nv] and PS[Np] coincide exactly.

I made a plot PL[Nv](H[Nv]) - Pres_isa(H[Nv]) (green color) and lev[Np](H[Np]) - Pres_isa(H[Np]) (blue color), where Pres_isa is the reference atmospheric pressure according to the International Standard Atmosphere ISO 2530 (1975). I subtracted the reference pressure model in order to see relatively small differences in plots. I made this computation for two sites: site #1: lat=60,lon=30 (St. Petersburg, at sea level) and site #2: lat= 35,lon=100 (Tibet, elevation around 5000 meters). Here the pressure anomalies Np Nv at site #1:
I computed spline coefficients for both PL[Nv](H[Nv]) and lev[Np](H[Np]) and re-gridded these functions to the same grid using spline interpolation. The differences Np minus Nv for site #1 are shown here:

Similar plots for site #2:
and

The differences Np minus Nv at some extent vary with time, but plots for different dates have the same feature: a minimum at 9000 m 90-10 Pa for site #1, a minimum at 11000 m 40-50 Pa for site #2.

Np Surface pressure

Summing all DELP and adding the Top of the Atmosphere pressure (1 Pa), I expected to get exactly the surface pressure PS. However, I found the difference 2.84 Pa for site #1 and 2.35 Pa for site #2.
This difference change with time:

Site #1  2011.09.20_00:00:00.0  Diff:   2.85 Pa
Site #1  2011.09.21_00:00:00.0  Diff:  -4.26 Pa
Site #1  2011.09.22_00:00:00.0  Diff:  -2.36 Pa
Site #1  2011.09.23_00:00:00.0  Diff:  -8.86 Pa
Site #1  2011.09.24_00:00:00.0  Diff:  -2.21 Pa
Site #1  2011.09.25_00:00:00.0  Diff:  -0.71 Pa
Site #1  2011.09.26_00:00:00.0  Diff:  15.95 Pa
Site #1  2011.09.27_00:00:00.0  Diff: -11.41 Pa

Back to Leonid Petrov's discussion page.
This page was prepared by Leonid Petrov (

)
2012.04.24_17:04:22