Determination of atmospheric temperature, water vapor, and heating rates from mid- and far- infrared hyperspectral measurements AGU Fall Meeting, Wednesday, December 12, 2007 GC34A 02 D.R. Feldman (Caltech); K.N. Liou (UCLA); Y.L. Yung (Caltech); D. G. Johnson (LaRC); M. L. Mlynczak (LaRC)
Presentation Outline Motivation for studying the far infrared FIRST instrument description Sensitivity i i tests of mid IR vs far IR capabilities i Clear sky Cloudy sky Multi instrument data comparison Climate model considerations Conclusions Outline 2
The Far Infrared Frontier Current EOS A-Train measure 3.4 to 15 μm, don t measure 15-100 μm IRIS-D measured to 25 μm in 1970 Far-IR, through H 2 O rotational band, affects OLR, tropospheric cooling rates Far-IR processes inferred from other spectral regions Mid-IR, Microwave, Vis/NIR Interaction between UT H 2 Oand cirrus clouds requires knowledge of both Currently inferred from measurements in other spectral regions Motivation No spectral measurements to the right of line Figures derived from Mlynczak et al, SPIE, 2002 3
FIRST: Far Infrared Spectroscopy of the Troposphere FTS w/ 0.6 cm 1 unapodized resolution, ±0.8 cm scan length Multilayer beamsplitter Germanium on polypropylene Good performance over broad spectral ranges in the far infrared 5 200 μm (50 2000 cm 1 1 ) spectral range NeDT goal ~0.2 K (10 60 μm), ~0.5 K (60 100 μm) ) 10 km IFOV, 10 multiplexed detectors Cooling Spectrometer LN 2 cooled Detectors liquid He cooled Scan time: 1.4 8.5 sec Balloon borne & ground based observations AIRS FIRST AIRS FIRST instrument 4
Retrieval Sensitivity Test Flow Chart T(z) H 2 O(z) O 3 (z) CWC(z) () CER(z) Model Atmosphere RTM + Noise Synthetic ti Measurement Random Perturbations A priori uncertainty A priori Atmospheric State) RTM A priori i spectrum Retrieval algorithm Sensitivity tests Analyze retrieved state, spectra, and associated statistics 5
Clear Sky Retrieval Test Sensitivity tests AIRS and FIRST T(z) retrievals comparable. FIRST better than AIRS in H 2 2O( O(z) retrievals eva 200-300 mbar. Residual signal in far IR seen 100-200 cm -1 low NeDT critical 6
Clear Sky Heating Rates Tropical Conditions Sub-Artic Winter Conditions Heating Rates Spectra provide information about fluxes/heating rates Error propagation (Taylor et al, 1994; Feldman et al, In Review) can be used Heating rate error for scenes with clouds generally higher h due to lack of vertical cloud information 7
Extrapolating Far IR with Clouds Retrieval of T(z), H 2 O(z), CWC(z), CER(z) difficult with AIRS spectra Use AIRS channels to extrapolate far IR channels? Depends on cloud conditions, T(Z), H 2 O(z) Low BT channels from 63μm 6.3 band low BT channels in far IR High BT channels scale from mid to far IR For tropics, channels with BT 250 270 K (emitting ~ 5 8 km) are complicated Clouds 8
Test Flight on September 18, 2006: Ft, Sumner NM AQUA MODIS L1B RGB Image FIRST Balloon AIRS Footprints Test flight CloudSat/CALIPSO Track 9
CloudSat/CALIPSO signals CloudSat and CALIPSO near collocation No signal from CloudSat CALIPSO signal consistent with FIRST residual ¾ Test flight 10
FIRST and AIRS Cloud Signatures Instrument collocation FIRST balloon-borne spectra AIRS MODIS Residuals are consistent with clouds ~ 5 km, D e ~ 60 μm Cloud Detected! Test flight 11
Climate Model Considerations Climate models produce fields that specify mid- & far-ir spectra. Multi-moment statistical comparisons of measured spectra and modeled spectra avoid subtle biases from data processing. Spectral and atmospheric state spaces should be considered jointly. Far-IR climate model analysis requires more far-ir data Far-IR extrapolation should retain physical basis and be verified with measurements. Agreement with CERES is only partial verification and presents a non-unique checksum Future work to assess how spectra impart information towards climate model processes. Model evaluation 12
Conclusions AIRS measures mid IR, but far IR is not covered A Train spectrometers. FIRST provides thorough h description of far IR but tlimited it spectra are available. FIRST clear sky T retrievals comparable, improved UT H 2O retrieval relative to AIRS Implied cooling rate information difference is small. Extrapolating far IR channels good for T b ~ 220 K, ok for T b ~ 300 K, difficult for T b ~250 270 K. Multi instrument analysis witha Train facilitates comprehensive understanding of FIRST test flight spectra. AIRS mid-ir spectra can validate climate models, but far-ir should not be neglected. Conclusions 13
Acknowledgements NASA Earth Systems Science Fellowship, grant number NNG05GP90H. Yuk Yung Radiation Group: Jack kmargolis, Vijay Nt Natraj, jking Fai FiLi Li, & Kuai ile George Aumann and Duane Waliser from JPL Xianglei Huang from U. Michigan and Yi Huang from Princeton AIRS, CloudSat, and CALIPSO Data Processing Teams Thank you for your time 14
Cloud Radiative Effect (CRE) CRE = TOA clear broadband flux TOA broadband dflux CERES provides collocated measurements of CRE from broadband radiometers Most CERES products contain multiple data streams AIRS L3 CRE lower than CERES CRE Other A Train sets (CloudSat/CALIPSO) can arbitrate difference Clouds 15
Towards CLARREO NRC Decadal Survey recommended CLARREO for Radiance calibration Climate monitoring CLARREO specified to cover 200 2000 cm -1 with < 2 cm -1 resolution NIST traceability requirement Prototyped far-ir instruments provide a science and engineering test-bed for next generation of satellite instruments Further orbital simulations required to test how mid-ir state space uncertainties appear as far-ir spectral residuals More integrated A-train analyses w.r.t. Far-IR warranted Larger Far-IR dataset analysis needed to demonstrate utility of long wavelength measurements for climate monitoring Don t forget about 50-200 cm -1 (200-50 μm). Future directions 16