ObsErrorFactorTopoRad

This obsfunction was designed to mimic GSI-Observer method of observation Error Inflation Factor (EIF).

Error Inflation Factor (EIF) as a function of terrain height, channel, and surface-to-space transmittance.

H = surface height [m]

X = surface-to-space transmittance

IASI, CrIS, AIRS and AVHRR3:

\(factor = (2000/H)^4\)

\(EIF = SQRT \{ 1 / [ 1 - (1 -factor ) * X] \}\) if \(H > 2000\), for all the input channels

AMSU-A:

\(EIF = SQRT [ 1 / ( 2000 / H ) ]\) if \(2000 < H < 4000\), for Channels 1-6,15

\(EIF = SQRT [ 1 / ( 4000 / H ) ]\) if \(H > 4000\), for Channel 7

ATMS:

\(EIF = SQRT [ 1 / ( 2000 / H ) ]\) if \(2000 < H < 4000\), for Channels 1-7,16

\(EIF = SQRT [ 1 / ( 4000 / H ) ]\) if \(H > 4000\), for Channel 8

Required input parameters:

Sensor

Sensor name: currently only works for these sensors iasi, cris-fsr, airs, avhrr3, amsua, atms

Channels

Used channels

Output parameters:

Error Inflation Factor (EIF) for each input channel of the input sensor at each location.

Required fields from obs/geoval:

geovals

GeoVaLs/surface_geopotential_height

obsDiag

ObsDiag/transmittances_of_atmosphere_layer

Example configurations:

- filter: BlackList
 filter variables:
 - name: brightnessTemperature
   channels: 1-15
 action:
   name: inflate error
   inflation variable:
     name: ObsFunction/ObsErrorFactorTopoRad
     channels: 1-15
     options:
       sensor: amsua_n19
       channels: 1-15

In this example, the observation errors of NOAA-19 AMSU-A are inflated according to the algorithm.