Observing methane from space Daniel J Jacob Observing

Observing methane from space Daniel J. Jacob

Observing system for atmospheric methane surface+aircraft monitoring Field campaigns Barnett Shale (EDF) Satellite (GOSAT) HIPPO

LEO (Low Earth Orbit) and GEO (Geostationary) LEO 500 -1000 km alitude Global observation Temporal frequency: up to 1 day (depends on cross-track scanning) GEO 36, 000 km altitude Regional-continental observation Temporal frequency: up to continuous

Observation modes from Low Earth Orbit

Observing methane from space in the shortwave infrared (SWIR) Atmospheric optical depths H 2 O H 2 O Io solar backscatter CH 4 IB CH 4 CO 2 CH CH 4 4 N 2 O N O CO 2 CO surface Backscattered radiation measures optical depth τ of atmosphere: where AMF is an air mass factor absorption x-section -2 Optical depth is related to columns Ω [molecules cm ]: Retrieve methane column ΩCH 4 by fitting the backscattered spectrum

SWIR retrievals at 1. 65 and 2. 3 µm SCIAMACHY GOSAT GHGSat GOSAT-2 Carbon. Sat TROPOMI GOSAT-2 GEO-CAPE geo. CARB H 2 O CO 2 CH 4 N 2 O CO CO 2 proxy method at 1. 65 µm from joint retrieval of methane and CO 2 columns: assumed known cancels effects of surface/atmospheric scattering and instrument biases

Observing methane from space in thermal infrared (TIR) GOSAT thermal IR spectrum CH 4 surface Herbin et al. [2013] Fitting of terrestrial radiation spectrum vertical concentration profile of methane Resolution limited by need for temperature contrast

Retrieval sensitivity in SWIR and TIR Column averaging prior kernel methane column [molecules cm-2] partial columns for different altitude levels Convert to dry-air column mean mole fraction: X = / a column of dry air Cloud-free conditions are required in both SWIR and TIR (elements of a) Worden et al. [2015]

Fraction of clear-sky scenes vs. satellite pixel size The likelihood of a clear scene (“cloud clearing”) is strongly dependent on satellite pixel resolution An instrument with 8 km pixels can retrieve successfully only ~10% of scenes Remer et al. [2012]

SWIR instruments for observing methane Instrument Low Earth Orbit Solar backscatter SCIAMACHY GOSAT TROPOMI GHGSat GOSAT-2 Carbon. Sat Active (lidar) MERLIN Geostationary GEO-CAPE geo. CARB Agency Data period Pixel size Coverage [km 2] Precision ESA JAXA ESA GHGSat, Inc. JAXA ESA 2003 -2012 200920162018 proposed 30 60 10 10 7 7 0. 05 x 0. 05 10 x 10 2 2 6 days 3 days (sparse) 1 day targets 3 days (sparse) 5 -10 days 1. 5 % 0. 7 % 0. 6% 1 -5% 0. 4% DLR/CNES 2020 - 50 50 monthly 1. 0% NASA proposed 4 4 4 5 hourly 2 -8 hours 1. 0%

Methane data from SCIAMACHY and GOSAT Dry-air column mean mole fractions XCH 4 Color scale adjusts for 30 ppb global increase in methane from 2003 -2004 to 2010 -2013

Characterization of measurement errors total error systematic bias (accuracy) random error (precision) • Random error decreases with sqrt(n) of independent observations (IID condition) • Global mean bias is not an issue because methane background is known • Regionally variable bias (“differential bias”) is irreducible and propagates to bias in inference of emissions

Direct validation using TCCON observation network methane column TCCON 1. 65 µm Precision Differential bias SCIAMACHY (2003 -2005) 30 ppb 4 -13 ppb GOSAT 12 -13 ppb 2 -3 ppb Needed for inversions 34 ppb 10 ppb TCCON measures the same column-mean mole fraction XCH 4 as the SWIR satellites Buchwitz et al. [2015]

Indirect validation using chemical transport models (CTMs) as common intercomparison platform CTM provides continuous 3 -D fields to compare different observational data sets GEOS-Chem CTM with prior emissions Satellite Diagnose satellite bias by inconsistency with suborbital data sets aircraft+surface data