TAFTS: CAVIAR field data from Camborne 2008 Ralph Beeby, Paul Green, Juliet Pickering, John Harries.

TAFTS: CAVIAR field data from Camborne 2008

Ralph Beeby, Paul Green, Juliet Pickering, John Harries

Contents

• Introduction• Camborne Field Campaign, Flight B400• Atmospheric Profiling• Uncertainty in Profiles• Spectra: LBLRTM simulation, TAFTS measurements,

intercomparison

Introduction

• TAFTS: Tropospheric Airborne Fourier Transform Spectrometer

• Far-Infrared FTS, designed and built at Imperial College

• Spectral range: 14-200µm (50-700cm-1)

• Resolution of 0.12cm-1 • Unique aircraft-mounted,

dual-input polarising FTS The TAFTS radiometer

Camborne Field Campaign

• First CAVIAR1 field measurement campaign

• Flights over Camborne, 13th August to 25th September 2008

• In-situ measurements of the infrared absorption spectrum of atmospheric water vapour

• TAFTS2: 14-200µm (50-700cm-

1)• and ARIES3: 3-25µm (600-

3000cm-1) – infrared interferometers mounted onboard FAAM4 BAe-146 research aircraft

• Auxiliary measurements of humidity (and other gas concentrations), temperature, pressure

1 Continuum Absorption in the Visible and Infrared and Its Atmospheric Relevance

2 Tropospheric Airborne Fourier Transform Spectrometer

3 Airborne Research Interferometer Evaluation System

4 Facility for Airborne Atmospheric Measurements

FAAM BAe-146, courtesy www.metoffice.gov.uk

Case Study: Flight B400

• Focussing on flight B400 – 18 September 2009

• Choice based on favourable weather conditions and instrument performance

• Auxiliary measurements: sondes used to measure humidity and temperature, plus in-situ measurements from aircraft

Methodology• Radiometers (TAFTS and ARIES) used to take radiance

measurements which are converted to absorption spectra• In order to make meaningful use of these spectra, it is necessary

to know the properties of the atmosphere above / below the aircraft – i.e., what is absorbing the incoming radiation?

• Use auxiliary measurements to estimate variation with altitude of humidity and temperature in the column of atmosphere between aircraft and top of atmosphere / ground.

• This profile of the atmosphere is used to simulate spectrum we would expect to see based on LBLRTM / HITRAN

• Compare LBLRTM output with TAFTS measurements – calibrate so monomer lines agree

• Study “window regions” in absorption spectra – absorption here is due to water vapour continuum

• Compare differences in window region absorption to estimate error in continuum

Atmospheric Profiling

• Need to know the distribution of water vapour above and below the aircraft in order to compare TAFTS measurements with LBLRTM1/HITRAN2

• Aircraft performs straight, level runs (SLRs) to take measurements• Collect data from different sources to produce a “best estimate” :• Dropsondes released from aircraft• Balloon radiosondes launched from Camborne 2-3 times daily• ECMWF3 forecast model• Collaboration with Stuart Newman (Met Office) and Liam Tallis (Reading) to find best scheme for determining profile

1 Line-by-line Radiative Transfer Model

2 High-resolution Transmission Molecular Absorption Database

3 European Centre for Medium-range Weather Forecasting

Vaisala RD93 dropsonde, courtesy www.vaisala.com

Atmospheric Profiling (2)

• Divide SLR into sections over land and over the sea

• Over the sea: use nearest dropsonde as most reliable source

• Use nearest radiosonde from top of dropsonde profile up to 100mb

• ECMWF up to top of model atmosphere

• Look at variation in ECMWF with time and space, apply similar variation to sonde profiles where necessary

Pre

ssure

/

mb

Water vapour / %RH

Temperature / K

Simulated Spectra

• LBLRTM Far-IR spectra for SLR at 34,000ft

• (Red line indicates Planck curve for 225.6K: UW spectrum above, DW below)

Wavenumber / cm-1

Radia

nce

/ m

W/m

2.s

r.cm

-1

Uncertainty in Profiles

• Uncertainty in atmospheric profiles is a source of uncertainty in simulated spectra – this affects comparison between simulations and TAFTS spectra

• Make use of Jacobians in LBLRTM to assess sensitivity of spectra to uncertainties in temperature and relative humidity

• Analytic Jacobians: calculate dR/dx across spectral range where R = radiance and x = atmospheric parameter

• Indicates the change in radiance that would be caused by a change in a given atmospheric parameter (e.g., temperature, relative humidity)

Jacobians for 34kft run

Wavenumber / cm-1

Radia

nce

/

mW

/m2.s

r.cm

-1Pre

ssu

re /

mb

dR

/d log

[vm

r(H

2O

)] /

m

Wm

2sr

.cm

-1/log

[vm

r]

Jacobians for 34kft run

dR

/d log

[vm

r(H

2O

)] /

m

Wm

2sr

.cm

-1/log

[vm

r]

Wavenumber / cm-1

Radia

nce

/

mW

/m2.s

r.cm

-1

TAFTS Spectra

• TAFTS spectrum for 34,000ft level run (TAFTS spectrum in blue)

• LW (50 – 300cm-1): comparison very promising between TAFTS and LBLRTM

• Not a perfect match -uncertainties due to– Noise and errors in TAFTS

measurements– Errors in RH profile– Errors in T profile– Error in water vapour

continuum• Once other uncertainties are

properly characterised, can estimate error in continuum

Wavenumber / cm-1

Radia

nce

/

mW

/m2.s

r.cm

-1

Coming Soon…

• More TAFTS spectra from other level runs• Noise and uncertainty analysis on spectra :

– Errors in atmospheric profiles• Uncertainty in RH• Uncertainty in T

– Jacobians indicate extent to which these errors affect simulated spectra

– Noise and error analysis of TAFTS measurements

• LBLRTM: repeat simulations using HITRAN 2008?• Estimate of continuum correction based on comparison

between measured and simulated spectra