Remote Sensing Using NASA EOS A-Train Measurements Presentation at Sonoma Technology, Inc. Monday, June 16, 2008 Daniel R. Feldman Caltech Department of.

Remote Sensing Using NASA EOS A-Train Measurements

Presentation at Sonoma Technology, Inc.Monday, June 16, 2008

Daniel R. Feldman Caltech

Department of Environmental Science and Engineering

Presentation Outline

• Overview of satellite-based remote sensing.

• Discussion of several EOS A-Train datasets.– AIRS, CloudSat, CALIPSO.

• Products derived from the datasets.– Standard retrieval products.

– Radiative heating/cooling rate profiles.

• The next generation of instrumentation.

• Conclusions.

2 OutlineOutline

The Power of Remote Sensing

• With measurements at different wavelengths:– Distribution of trace gases.– Aerosols and cloud properties.– Energy balance/exchange.

• From satellite-based measurements, we obtain a comprehensive, quantitative picture used to (in)validate earth science hypotheses.

• Measurements have implications for policy.

Remote Sensing & SocietyRemote Sensing & Society

The EOS A-Train Data Age

• The polar-orbiting EOS A-Train flotilla presents a voluminous dataset describing the earth’s lower atmosphere:

– Aqua platform operational for ~ 6 years.

– CloudSat and CALIPSO platforms operational for ~ 2 years.

• This data can be very scientifically useful in the context of measurement/ model comparisons.

4 DatasetsDatasets

Artist’s rendition of the A-Train courtesy of NASA

Dataset Overview

• Many disparate datasets measuring at different wavelengths.– AIRS: hyperspectral, cross-track scanning mid-IR data.

• T profiles within 1 K/km, H2O profiles within 15 % / 2km.

• Near-global coverage on a daily basis.

– CloudSat/CALIPSO: cloud water content profiles from radar/lidar.

• 50% CWC uncertainty / 240 m.

• Near-global coverage on a bi-weekly basis.

– Other instruments in the A-Train shed light on current earth science questions.

5 DatasetsDatasets

AIRS Instrument• Grating spectrometer measures

3.7 to 15.4 μm (650-2700 cm-1).• Cross-track scanning mirror

yields 90 footprints in 2.7 sec.• Space & BB view for calibration.• Each footprint produces 2378

radiance measurements..• 15 km footprint.• Collocated 15-channel passive

microwave sounder at 45 km footprint.

From JPL AIRS website

DatasetsDatasets

AIRS Achievements

• Unprecedented view of temperature, water vapor, and carbon dioxide distribution on a bi-weekly basis.

7

Avg Trop Relative Humidity From AIRS, Dec-Feb 2002-2005

DatasetsDatasets

CloudSat Overview• CloudSat

– Nadir-pointing 94-GHz radar– Cloud-profiles at ~240 m

vertical resolution – Horizontal resolution ~1.4 km – Sensitivity of -31 dBZ, 80 dBZ

dynamic range

Horiz. Res.

Vert. Res.

DatasetsDatasets

CALIPSO Overview• CALIPSO: Cloud-Aerosol LIdar with Orthogonal

Polarization– Nadir-pointing 2-channel (532 nm and 1064 nm) lidar.– Vertical resolution ~30 m.– Horizontal resolution ~100 m.– Min τvis sensitivity of 0.005, max τvis = 5.

• Combined product with CALIPSO offers detailed understanding of cloud vertical distribution

heig

ht (

km,

MS

L) cloudsat

calipso

DatasetsDatasets

CloudSat/CALIPSO Achievements

10 DatasetsDatasets

• Unprecedented global coverage of cloud-profile distribution on a seasonal basis.

JJA zonally averaged distribution of cloudiness derived from the CloudSat 2B-GEOPROF product.

JJA zonally averaged distribution of cloudiness from one of the IPCC FAR climate models , from Mace and Klein.

Interpreting Measurements

• Raw measurements are inverted into higher level products.

• Inversion requires understanding of radiative transfer.– Planck emission.

– Absorption features: line strengths, broadening/continuum.

– Optical properties of scatterers.

– Mechanics of integrating fundamental eqn. of RT.

From JARS RT tutorial

From Goody & Yung, Ch 1

InversionInversion

Inversion of Measurements

• With a working RT model, profile quantities can be derived from the measurements.

• However, problem is ill-conditioned => methods required to produce mathematical stability.

From Boesch, et al, 2006

InversionInversion

Derivation of Retrieval Products

• NASA satellite instrument data processing protocols specify several levels of products:– L1A: raw measurements

– L1B: geolocated, calibrated measurements

– L2: retrieved from L1B data, forward model, etc.

– L3: gridded, averaged L2 products

• Higher-level products should be utilized with care– Meaningful scientific analysis requires full tabulation of

the retrieval deficiencies.

13 InversionInversion

Circulation Models & Radiation

14

Predict T, q, u

PBL & Surface

Radiation

Dissipation Terms

Solution of Primitive Equations

Prediction of Condensation

Cloud Fraction

• Stratosphere in approximate radiative equilibrium → SW heating ≈ IR cooling.

• In troposphere, IR cooling>SW heating.

• Circulation model performance requires proper treatment of radiative energy exchange.

Flowchart of model calculation for an isolated timestep from Kiehl, Ch. 10 of Trenberth, 1992

Novel productsNovel products

Cooling Rate Profile Uncertainty

• Perturbations in T, H2O, O3 profiles lead to θ’ changes that propagate across layers.

• Calculation of θ’ uncertainty requires formal error propagation analysis.

15

n

i

n

jji

ji

xxx

z

x

zz

1 1

2 ,cov

From Feldman, et al., 2008.

Novel productsNovel products

Retrieval of Cooling Rates

• Many products derived from the satellite instrument measurements through retrievals.

• Many different approaches to retrieving quantities from measurements.

16

From Feldman, et al., 2006.

Novel productsNovel products

CloudSat Heating/Cooling Rates

17From Feldman, et al., In Review

• Radar reflectivity → CWC profiles + ECMWF T, H2O, O3 → fluxes and heating rate profiles (2B-FLXHR).

• Uncertainty estimates not given in current (R04) release.

Novel productsNovel products

Net Heating from CloudSat/CALIPSO

18From Feldman, et al., In Review Novel productsNovel products

Moving from OLR to Cooling Rates

19

• Qualitative agreement between measurement/model mean OLR values

• Different cooling rate profiles, though OLR, cooling rates are closely related.

From Feldman, 2008

Novel productsNovel products

CLARREO: The Next Generation

20

• Fundamental differences between measurements and climate models and in key feedback descriptors for IPCC FAR models.

• Long-term trend characterization & attribution from satellite instruments is very difficult.

– NRC 2007 Decadal Survey recommended the development of an instrument that is NIST-calibrated in orbit.

• CLimate Absolute Radiance and Refractivity Observatory (CLARREO) will have high spectral resolution in the visible, mid- and far-IR.

Future missions Future missions

FIRST: Far Infrared Spectroscopy of the Troposphere

• FIRST is a test-bed for CLARREO

• NASA IIP FTS w/ 0.6 cm-1 unapodized resolution, ±0.8 cm scan length

• 5-200 μm (2000 – 50 cm-1) spectral range

• NeDT goal ~0.2 K (10-60 μm), ~0.5 K (60-100 μm)

• 10 km IFOV, 10 multiplexed detectors

• Balloon-borne & ground-based observations

21

FIRSTAIRS AIRS

Future missions Future missions

Towards CLARREO

• CLARREO, as a future NASA mission, is currently being studied by several institutions.– Exacting engineering requirements to achieve NIST calibration.

• Test-bed instrumentation under development– FIRST provides a comprehensive description of the far-infrared which

is relevant to CLARREO development.

• Establishing climate trends from satellite data and attributing causes to these trends is within reach.– With the establishment of a benchmark, climate model discrepancies

can be rectified.

22 Future missions Future missions

Conclusions• Satellite-based remote sensing is a powerful tool for earth

science.• Proven utility to society for nearly almost 40 years.

• EOS A-Train data contain information about many aspects of the earth-atmosphere system:• Temperature profile, trace gas constituents, cloud profiles.• Description of fields that are of direct relevance to weather and

climate model evaluation (e.g., radiative energy exchange).

• The next generation of satellite instruments will be designed not just for process and trend description.• Climate models will directly motivate mission specifications.

23 ConclusionsConclusions

Acknowledgements

• NASA Earth Systems Science Fellowship, grant number NNG05GP90H.

• Yuk Yung Radiation Group: Jack Margolis, Vijay Natraj, King-Fai Li, & Kuai Le, Xi Zhang, Xin Guo

• George Aumann, Duane Waliser, Jonathan Jiang, and Hui Su from JPL.

• Tristan L’Ecuyer from CSU.

• Marty Mlynczak and Dave Johnson of NASA LaRC.

• Xianglei Huang from U. Michigan.

• Yi Huang from Princeton.

• AIRS, CloudSat, and CALIPSO Data Processing Teams.

24 Thank you for your timeThank you for your time