Meteorologist, oceanographer, Python solvem probler, experimenter, Raspberry Pi tinkerer, geek dad, manager, space enthusiast, ham radio lover, science columnist, ADHDer, synthetic intellect curious, linux admin cosplayer.

Post 48 Research — Satellite Radiance Instruments & Operational-Adoption Timeline

STATUS: COMPLETE (research-agent dossier for “The Hemisphere That Caught Up”) Scope: instruments + operational-adoption dates, channel counts, orbit types, sources. Convention: dates/counts VERIFIED against fetched sources unless marked FLAG / NOT VERIFIED.

THE PAYOFF (frame / verification anchor)

Simmons, A.J. & Hollingsworth, A. (2002), “Some aspects of the improvement in skill of numerical weather prediction.” Q.J.R. Meteorol. Soc. 128: 647–677. DOI 10.1256/003590002321042135. (Also ECMWF Tech Memo 342.)

  • Documents the ~1-day gain in 500 hPa height / MSLP predictability over the preceding decade in the NH, and “a similar gain over the last 3 years in the southern hemisphere.”
  • Direct quote (payoff line): “One-day forecast errors have been reduced so much in the southern hemisphere that medium-range forecasts for the region have become almost as skilful as those for the northern hemisphere.”
  • This is the SAME verification framework cited in Post 47. The SH/NH anomaly-correlation convergence is the canonical chart. Agent of convergence = satellite radiance assimilation.
  • Source: https://rmets.onlinelibrary.wiley.com/doi/10.1256/003590002321042135 ; ECMWF TM342: https://www.ecmwf.int/sites/default/files/elibrary/2001/12238-some-aspects-improvement-skill-numerical-weather-prediction.pdf

INSTRUMENT CHRONOLOGY (the scaffold)

1. VTPR — Vertical Temperature Profile Radiometer (1972)

  • 8-channel infrared sounder (7 channels in the 15-um CO2 band + 1 window channel). CONFIRM exact split; NCEI calls it an “operational 8-channel sounding system.”
  • Flown on NOAA-2 through NOAA-5 (ITOS = Improved TIROS Operational System).
  • NOAA-2 (ITOS-D) launched 15 October 1972. Routine soundings Nov 1972 – Feb 1979.
  • Orbit: polar sun-synchronous. Horizontal resolution ~55x57 km at nadir.
  • First operational thermodynamic soundings from a NOAA spacecraft; direct predecessor of HIRS.
  • Larry McMillin (NOAA/NESDIS) authored the definitive early VTPR technical report.
  • Sources: NCEI VTPR https://www.ncei.noaa.gov/products/vertical-temperature-profile-radiometer ; McMillin NOAA-TR-NESS65 https://www.ncei.noaa.gov/sites/default/files/2021-08/noaa-tr-ness65-McMillin-VTPR.pdf

2. TOVS on TIROS-N (1978) = HIRS/2 + MSU + SSU

  • TIROS-N launched 13 October 1978. Orbit: sun-synchronous ~833/870 km, equator crossings 0730 (AM) / 1430 (PM). Carried forward on NOAA-6..14.
  • HIRS/2 (High Resolution Infrared Radiation Sounder): 20 channels — 19 IR + 1 visible (0.70 um). Includes the 15-um CO2 band (ch 1–5), 11.1-um window (ch 8), 9.7-um ozone (ch 9), 6-um water-vapor bands (ch 10–12), and 4-um bands.
  • MSU (Microwave Sounding Unit): 4 channels — one 50.31 GHz window + three in the 55 GHz O2 band (53.73, 54.96, 57.95 GHz). Temperature; cloud-insensitive.
  • SSU (Stratospheric Sounding Unit): 3 channels at 15.0 um via pressure-modulated CO2 cells (upper-stratosphere temperature).
  • This is McMillin’s split-window / retrieval (TOVS) era: radiances INVERTED into temperature profiles, then the profiles assimilated.
  • Sources: CEDA TOVS record https://catalogue.ceda.ac.uk/uuid/100dacbe559b4004a6d024cbea0aa617/ ; NOAA/NESDIS CLASS TOVS/ATOVS https://www.class.noaa.gov/release/data_available/tovs_atovs/index.htm

3. ATOVS on NOAA-15 (1998) = AMSU-A + AMSU-B + HIRS/3

  • NOAA-15 launched 13 May 1998 — first spacecraft to fly AMSU. Polar sun-synchronous.
  • AMSU-A: 15 channels between 23.8 and 89 GHz — atmospheric temperature sounding; replaced MSU (4 ch) + SSU (3 ch) with far finer vertical resolution. The single most important microwave temperature sounder.
  • AMSU-B: 5 channels between 89 and 183.3 GHz — moisture sounding; ~15 km resolution near nadir. (Succeeded by MHS on NOAA-18 onward.)
  • HIRS/3 IR sounder also aboard. Together the three = ATOVS (Advanced TIROS Operational Vertical Sounder).
  • Source: Wikipedia AMSU https://en.wikipedia.org/wiki/Advanced_microwave_sounding_unit ; NASA Aqua AMSU https://aqua.nasa.gov/content/amsu

4. ATMS on Suomi-NPP (2011) — successor microwave sounder

  • Suomi-NPP launched 28 October 2011 (FLAG — date from general knowledge, not re-fetched). Polar sun-synchronous.
  • ATMS (Advanced Technology Microwave Sounder): 22 channels combining AMSU-A + AMSU-B/MHS heritage into one instrument. NOT VERIFIED against a fetched primary source — confirm channel count before use.

5. AIRS on Aqua (2002) — hyperspectral IR (grating spectrometer)

  • Aqua/AIRS launched 4 May 2002. Sun-synchronous (A-train, 1330 ascending).
  • 2378 infrared channels, 3.7–15.4 um; 13.5 km footprint. Flown with AMSU-A on Aqua.
  • ~100x the spectral resolution of prior IR sounders — first hyperspectral IR sounder assimilated operationally.
  • Source: Wikipedia AIRS https://en.wikipedia.org/wiki/Atmospheric_infrared_sounder ; NASA JPL AIRS https://airs.jpl.nasa.gov/mission/overview/

6. IASI on MetOp-A (2006) — hyperspectral IR Fourier-transform interferometer

  • MetOp-A launched 19 October 2006 (Baikonur). Polar sun-synchronous (0930 descending).
  • Michelson / Fourier-transform interferometer. Spectral coverage 645–2760 cm^-1 (15.5–3.62 um); spectral sampling 0.25 cm^-1, apodized resolution 0.5 cm^-1.
  • 8461 spectral samples (channels) — VERIFIED via Wikipedia IASI (“8461 spectral samples aligned in 3 bands”). Matches the remit’s ~8461.
  • Source: Wikipedia IASI https://en.wikipedia.org/wiki/Infrared_atmospheric_sounding_interferometer

7. SSM/I on DMSP (from 1987) — microwave imager (context)

  • DMSP F8 launched 18 June 1987 — first SSM/I; first microwave imager sensor. Polar sun-synchronous (US Air Force DMSP).
  • 7 channels (dual-polarized) at 19.35, 22.235, 37.0, 85.5 GHz. Products: surface wind speed, total column water vapor, cloud liquid water, rain rate. Feeds NWP over data-sparse oceans.
  • Source: Wikipedia SSM/I https://en.wikipedia.org/wiki/Special_sensor_microwave/imager ; NCEI DMSP microwave imager https://www.ncei.noaa.gov/products/dmsp-microwave-imager

8. Geostationary sounders/imagers (context)

  • Meteosat / GOES imagers supply water-vapor-channel radiances and atmospheric motion vectors (AMVs). Geostationary orbit (~35786 km), fixed sub-satellite point. Used for context vs polar sounders; not the SH-catch-up driver (geostationary coverage thins toward the poles).

9. GPS radio occultation (bias-free vertical profiles)

  • GPS/MET (1995) — proof-of-concept experiment (aboard MicroLab-1); demonstrated RO retrievals of temperature/refractivity. LEO orbit.
  • CHAMP (2000) and SAC-C — follow-on single-satellite RO. LEO.
  • COSMIC / FORMOSAT-3 — six-microsat constellation launched 15 April 2006 (Vandenberg); first operational GPS RO mission (US–Taiwan).
  • Technique: measure phase/amplitude of occulted GPS signals -> bending angle -> refractivity -> temperature/pressure/water vapor. High vertical resolution; essentially bias-free and self-calibrating (no radiative-transfer model, needs no bias correction) — hence anchors the radiance bias corrections.
  • Human centres: Rick Anthes, Christian Rocken (UCAR/COSMIC); Sean Healy (ECMWF RO assimilation).
  • Source: COSMIC/FORMOSAT-3 BAMS 2020 https://journals.ametsoc.org/view/journals/bams/101/7/bamsD180290.xml

THE ENGINE: retrieval -> direct radiance assimilation (software)

  • Eyre 1989 (QJRMS): “Inversion of cloudy satellite sounding radiances by nonlinear optimal estimation. I: Theory and simulation for TOVS.” — the optimal-estimation framework.
  • Eyre 1991, ECMWF Tech Memo No. 176: “A fast radiative transfer model for satellite sounding systems.” — origin of RTTOV (RTTOV-3 used operationally at ECMWF 1992–1998). RTTOV = the fast forward model H that makes direct radiance assimilation tractable (millions of evaluations per cycle).
  • Eyre et al. 1993 (QJRMS 119): “Assimilation of TOVS radiance information through one-dimensional variational analysis” — 1D-Var, the bridge from retrieval to variational.
  • RTTOV now maintained through the EUMETSAT NWP SAF (Roger Saunders). Fast-RT model paper: NWP SAF Tech Memo 282.
  • The turn: instead of inverting radiances to profiles first, put the radiative-transfer forward model INSIDE the variational cost function as the observation operator H and assimilate raw brightness temperatures directly. Direct radiance assimilation is the first great customer of variational DA (3D/4D-Var, Post 47).
  • Sources: Eyre 1993 QJRMS https://rmets.onlinelibrary.wiley.com/doi/abs/10.1002/qj.49711951411 ; NWP SAF RTTOV TM282 https://nwpsaf.eu/oldsite/deliverables/rtm/papers/tm282.pdf

Variational bias correction (VarBC)

  • Dee 2004, ECMWF Workshop on Assimilation of High Spectral Resolution Sounders: “Variational bias correction of radiance data in the ECMWF system.” https://www.ecmwf.int/sites/default/files/elibrary/2004/8930-variational-bias-correction-radiance-data-ecmwf-system.pdf
  • Auligne, McNally & Dee 2007 (QJRMS 133: 631–642): “Adaptive bias correction for satellite data in a numerical weather prediction system.” https://rmets.onlinelibrary.wiley.com/doi/abs/10.1002/qj.57
  • Bias correction coefficients estimated inside the analysis; anchored by bias-free data (RO, radiosondes). Essential to making millions of radiances assimilable.

Observation impact / FSOI (the proof AMSU-A is king)

  • Langland & Baker 2004 (Tellus A): adjoint-based observation-impact (FSOI) methodology.
  • Cardinali 2009 (QJRMS; ECMWF FSO tech report): FSO as an operational diagnostic. https://www.ecmwf.int/sites/default/files/elibrary/2009/8578-forecast-sensitivity-observations-fso-diagnostic-tool-monitoring-impact-observations-short.pdf
  • Consistent finding across centres: AMSU-A delivers the largest total forecast-error reduction of any observing system, followed by IASI, then radiosondes (TEMP), aircraft, SYNOP. Data-denial / OSE experiments show SH skill collapses toward ~1980s levels without satellites.
  • Nancy Baker, Rolf Langland (NRL); Carla Cardinali (ECMWF); Sanghyun Joo — FSOI human centres.

OPERATIONAL-ADOPTION TIMELINE (by centre)

Centre Milestone Date Note
NOAA/NESDIS VTPR operational soundings Nov 1972 retrieval era begins
NOAA/NESDIS TOVS (HIRS/2+MSU+SSU) on TIROS-N 13 Oct 1978 retrieval / McMillin era
NCEP SSI (3D-Var) operational June 1991 first operational 3D-Var (GDAS/GFS)
ECMWF 3D-Var operational (direct TOVS cloud-cleared radiances) 30 Jan 1996 ran to 24 Nov 1997
ECMWF 4D-Var operational 25 Nov 1997 Post 47 spine
NCEP direct IR+MW radiance assimilation in GFS (clear-sky) 1998 Derber & Wu 1998
NOAA NOAA-15 / first AMSU launched 13 May 1998 ATOVS era
ECMWF ATOVS (incl. AMSU-A) radiances operational May 1999 after 10 Mar 1999 TL319L50 upgrade; McNally et al. 1999
Met Office 3D-Var operational (Unified Model) ~1999 FLAG — Lorenc et al. 2000; date approximate, not re-fetched
JMA 1D-Var RTOVS (NOAA-14) in global analysis 23 Mar 2000 JMA radiance start; later ATOVS direct radiance
NASA AIRS on Aqua launched 4 May 2002 hyperspectral IR
EUMETSAT IASI on MetOp-A launched 19 Oct 2006 ~8461 channels
US/Taiwan COSMIC/FORMOSAT-3 (GPS RO) launched 15 Apr 2006 first operational RO constellation
NPP ATMS on Suomi-NPP launched 28 Oct 2011 FLAG — date general knowledge

Key ECMWF human centres (direct radiance in variational DA): Erik Andersson, Anthony (Tony) McNally, Jean-Noel Thepaut, Graeme Kelly, Jean Pailleux. Andersson et al. 1994 developed the methods for direct TOVS radiance use; the full 3D-Var implementation trilogy is Courtier et al. 1998 (I: Formulation), Rabier et al. 1998 (II), Andersson et al. 1998 (III: Experimental results).

Source (ECMWF 3D-Var operational date + ATOVS May 1999): ECMWF IFS documentation / Courtier et al. 1998 https://rmets.onlinelibrary.wiley.com/doi/abs/10.1002/qj.49712455002 ; NCEP data assimilation history https://www.emc.ncep.noaa.gov/emc/pages/numerical_forecast_systems/ncep_data_assimilation.php

FLAGS / NOT VERIFIED (handle with care before publishing)

  • VTPR exact channel breakdown (7 CO2 + 1 window) — “8-channel” confirmed; per-channel split CONFIRM.
  • ATMS on Suomi-NPP: 22 channels and 28 Oct 2011 launch — NOT re-fetched from primary source.
  • Met Office 3D-Var operational year (~1999, Lorenc et al. 2000) — approximate; verify exact date.
  • IASI apodized resolution quoted as 0.5 cm^-1 (0.25 cm^-1 sampling) — both figures appear in Wikipedia; standard EUMETSAT figures, but cross-check against EUMETSAT primary if precision matters.
  • Simmons & Hollingsworth “3 years” SH gain phrasing is from the 2002 paper’s abstract as summarized — verify exact wording against the PDF before quoting verbatim.

SOURCES (URLs actually fetched or returned in searches this session)

  • Simmons & Hollingsworth 2002 QJRMS: https://rmets.onlinelibrary.wiley.com/doi/10.1256/003590002321042135
  • ECMWF TM342 (S&H 2002): https://www.ecmwf.int/sites/default/files/elibrary/2001/12238-some-aspects-improvement-skill-numerical-weather-prediction.pdf
  • NCEI VTPR: https://www.ncei.noaa.gov/products/vertical-temperature-profile-radiometer
  • McMillin VTPR NOAA-TR-NESS65: https://www.ncei.noaa.gov/sites/default/files/2021-08/noaa-tr-ness65-McMillin-VTPR.pdf
  • CEDA TOVS instrument record: https://catalogue.ceda.ac.uk/uuid/100dacbe559b4004a6d024cbea0aa617/
  • NOAA CLASS TOVS/ATOVS: https://www.class.noaa.gov/release/data_available/tovs_atovs/index.htm
  • Wikipedia AMSU: https://en.wikipedia.org/wiki/Advanced_microwave_sounding_unit
  • NASA Aqua AMSU: https://aqua.nasa.gov/content/amsu
  • Wikipedia AIRS: https://en.wikipedia.org/wiki/Atmospheric_infrared_sounder
  • NASA JPL AIRS overview: https://airs.jpl.nasa.gov/mission/overview/
  • Wikipedia IASI: https://en.wikipedia.org/wiki/Infrared_atmospheric_sounding_interferometer
  • Wikipedia SSM/I: https://en.wikipedia.org/wiki/Special_sensor_microwave/imager
  • NCEI DMSP microwave imager: https://www.ncei.noaa.gov/products/dmsp-microwave-imager
  • COSMIC/FORMOSAT-3 BAMS 2020: https://journals.ametsoc.org/view/journals/bams/101/7/bamsD180290.xml
  • Eyre 1993 QJRMS 1D-Var: https://rmets.onlinelibrary.wiley.com/doi/abs/10.1002/qj.49711951411
  • NWP SAF RTTOV TM282: https://nwpsaf.eu/oldsite/deliverables/rtm/papers/tm282.pdf
  • Dee 2004 VarBC (ECMWF): https://www.ecmwf.int/sites/default/files/elibrary/2004/8930-variational-bias-correction-radiance-data-ecmwf-system.pdf
  • Auligne, McNally & Dee 2007 QJRMS: https://rmets.onlinelibrary.wiley.com/doi/abs/10.1002/qj.57
  • Cardinali FSO (ECMWF 2009): https://www.ecmwf.int/sites/default/files/elibrary/2009/8578-forecast-sensitivity-observations-fso-diagnostic-tool-monitoring-impact-observations-short.pdf
  • Courtier et al. 1998 3D-Var I (QJRMS): https://rmets.onlinelibrary.wiley.com/doi/abs/10.1002/qj.49712455002
  • NCEP data assimilation history: https://www.emc.ncep.noaa.gov/emc/pages/numerical_forecast_systems/ncep_data_assimilation.php