Summary of GEO-CAPE Aerosol Working Group Progress Members: Mian Chin, Pubu Ciren, Hiren Jethva, Joanna Joiner, Shobha Kondragunta, Alexei Lyapustin, Shana Mattoo, Lorraine Remer, Omar Torres, Alexander Vassilkov, Jun Wang

GEO-CAPE Aerosol Working Group tasks, FY13 TaskPersonnel High priority tasks: H1. Availability of aerosol retrieval at TEMPO pixel resolution with 1- km cloud masks Lorraine Remer, Shana Mattoo H2. Aerosol retrieval availability and product quality at TEMPO vs. GEO-CAPE pixel resolution with the cloud mask at the same pixel resolution Omar Torres, Hiren Jethva, Shana Mattoo H3. Feasibility of TEMPO-GOES-R synergistic retrieval: (a) expanded spectral range from UV (TEMPO) and IR (GOES-R), (b) with high resolution cloud mask (1-km GOES-R) for TEMPO, and (c) feasibility of two-angle retrieval (1)NESDIS group: Shobha Kondragunta, Pubu Ciren (2)GSFC group: Alexei Lyapustin (3)UNL group: Jun wang H4. Feasibility of retrieving aerosol layer height with O 2 -B and O 2 -  bands Alexander Vasilkov, Joanna Joiner Medium priority tasks: M1. Cloud contamination statistics for a variety of possible sensor resolutions, including TEMPO and GEO-CAPE Lorraine Remer, Shana Mattoo M2. Using 7-month GLI data to explore the GEO-CAPE capability: (a) build a database of surface reflectance at 380 nm, (b) retrieve SSA and aerosol height, and (c) identify aerosol types Alexei Lyapustin, Omar Torres

High priority tasks Can TEMPO achieve GEO-CAPE objectives by itself or by TEMPO/GOES-R synergy? – Aerosol retrieval availability: TEMPO will have 30-40% less retrieval availability then GEO-CAPE, unless the cloud screen criteria is much relaxed, which may increase AOD retrieval error – Aerosol product quality: not clear, still under investigation – Using GOES-R 2.1 μm channel to help TEMPO retrieve AOD at visible wavelength: Yes under certain conditions How much can be gained from 2-angle retrieval (GOES-R and TEMPO or GEO-CAPE)? – Combination of UV and Vis-IR provide more information for aerosol retrieval Can O 2 -B (688 nm) or O 2 -  (628 nm) band be used to retrieval aerosol height? – Yes with O 2 -B, when AOD is above 0.2, SSA and surface reflectance are known

GEO-CAPE vs. TEMPO GEO-CAPETEMPO Instrument resolutionBetter than 2x2 km, with 1x1 km cloud camera (2 spectral bands) 4.5x2 km Aerosol product resolution8x8 km (64-pixel)9x8 km (8-pixel) Temporal resolution1 hour AOD precision0.05 AAOD precision Spectral rangeUV-Vis (+ NIR, CO channel)UV-Vis AOCHYesNo

H1. TEMPO aerosol retrieval availability - Using MODIS cloud mask at 0.5 km (Lorraine Remer, Shana Mattoo) Regional statistics, 1x1 km (GEO-CAPE) and 4x2 km (TEMPO):  TEMPO would have about 55% of the retrievals that MODIS provides now  TEMPO will have 2/3 retrieval availability of GEO- CAPE (1x1 km)

H1’. Diurnal variation of “cloud-free” fraction with different pixel size  Cloud-free fraction can be significantly reduced at TEMPO pixel resolution especially in mid-day Virginia (VA) Wyoming (WY) New Mexico (NM) Mexico (MX) 8-km 4-km 2-km 1-km Diurnal variation of CLOUD-FREE fraction from 1-day data of August 12, 2010 Terra Aqua 4x2 km

H2. Aerosol retrieval availability and product quality at TEMPO and GEO-CAPE resolution (own cloud mask) (Omar Torres, Hiren Jethva)  Using the same variability criteria, TEMPO would have only half coverage of GEO- CAPE on average with the most severe loss in the SW largely because of the variation of surface reflectance. The availability greatly increases with the relaxed criteria Fraction clear sky 1x1 km (GEO-CAPE) Fraction clear sky 4x2 km (TEMPO) σ < Fraction clear sky 4x2 km (TEMPO) σ < Fraction clear sky 4x2 km (TEMPO) σ < 0.01 σ < abc

H2. Results – AOD relative quality σ=0.0025σ=0.005σ=0.01 Comparisons between 8x8 km AOD product from “GEO-CAPE” 1x1 km (x-axis) retrieval and “TEMPO” 4x2 km (y-axis) retrieval with different cloud mask criteria for April 10, 2012:  From the preliminary results, it does not seem that the quality of AOD from more relaxed cloud mask (thus higher probability of cloud contamination) is worse than that from more strict cloud mask. More quantitative evaluation is in progress. abc AOD 470 nm, 1x1 km AOD 470 nm, 4x2 km

H3. TEMPO/GOES-R Synergy – (Shobha Kondragunta, Pubu Ciren) (a) Surface reflectance at 2.13 μm and cloud mask GOES-R GOES-R remapped to TEMPO TEMPO  GOES-R (west) and TEMPO do not overlap over eastern part of US and Atlantic Ocean.  Synergic use of GOES-R is possible by providing surface reflectance at 2.13 μm for TEMPO aerosol retrieval over portion of eastern, middle and western part of U.S.  Observations in the early morning and later afternoon need to be avoided. Only hours close to noon time are favorable (for lambertian surface) July 15, 2012, UTC=19:00

H3. TEMPO/GOES-R Synergy – NESDIS group: (b) GOES-R cloud mask for TEMPO MODIS 1Km cloud mask (MOD35) was mapped into GOES-R (West) first then remapped GOES-R to TEMPO Grid between latitude band (35°N to 45°N) Fraction of cloudy (confident and probably cloudy) pixels calculated GOES-R remapped to TEMPO TEMPO  Cloud mask information from GOES-R observations seems to be sufficient for TEMPO July 15, 2012

H3. TEMPO/GOES-R Synergy (Alexei Lyapustin) Summer, July 11 Winter, December 2 120°W 105°W 75°W 120°W 105°W 75°W SNR=10 SNR=30 SNR=10 SNR=30  Over dark surfaces, the 2.1  m BRDF from GOES-R (East) model can be used for TEMPO aerosol retrieval regardless of TEMPO’s parking longitude  Over central-eastern USA, the AOT RMSE based on GOES-R BRDF is expected to be < 0.05 (at 0.47  m). Over western USA the error is larger

H3. TEMPO/GOES-R Synergy (Jun Wang) Angular Analysis of Vis/NIR sfc. Ratio and implication for retrieving AOD GOES-R 470, 640, 860 nm TEMPO/GEO-CAPE 340, 380, 470, 640 nm GOES-R + TEMPO/GEO-CAPE VZA (°) VZA (°) VZA (°)  azimuth  Azimuth  Azimuth  Azimuth  azimuth  Azimuth Degree of Freedom for retrieving AOD  Assuming that surface BRDF can be characterized with uncertainty of 20%, we found that strong synergy between two GOES-R and TEMPO/GEO-CAPE during the time when one of the sensor is viewing the Earth surface at the direction close to the Sun’s illumination angle.  TEMPO/GEO-CAPE UV wavelengths improves DOFs in AOD retrieval. SZA = 40°

H4. Aerosol height from O 2 -B and O 2 -  (Alexander Vasilkov, Joanna Joiner) Radiative transfer simulations – LIDORT code: line-by-line O 2 absorption,), TOA radiance with Gaussian response, FWHM=0.5, 1.0 – Aerosol: 1-km thick layer at different plume top height of 1-10 km, AOD=0.2, 0.5, 1.0, SSA=0.8, 0.9, 1.0 – Surface albedo: A=0.05, 0.1 Simulated TOA spectral reflectivity B-band ✔ SSA=1  If SSA is known, the aerosol plume height and optical depth can be derived from measurements of the band depth and continuum reflectivity at 686 nm

Conclusions Spatially, TEMPO alone (4x2 km) will have lower retrieval availability then GEO-CAPE (1x1 km) by 33% in the morning (10:30 am) Temporally, TEMPO will have reduced AOD retrieval availability in partial cloudy scene AOD quality from TEMPO and GEO-CAPE is still under evaluation Synergy between TEMPO and GOES-R is very helpful in terms of utilizing the extended spectral range for (1) retrieving AOD at visible wavelength and (2) increasing the degree of freedom for AOD and fine mode AOD retrievals O 2 -B band has promising sensitivity for aerosol height retrieval for both GEO-CAPE and TEMPO Bottom line: GEO-CAPE measurement requirements, as defined in STM and studied by the AeroWG, are not 100% met by TEMPO alone.

However… GEO-CAPE objectives can be mostly met by the synergy among TEMPO and GOES-R – GOES-R: high pixel res, cloud mask, DT retrieval, good for DT and ocean – TEMPO: bright surface, SSA – Combo: more aerosol retrieval Cloud screening: 1-km, 2-channel radiometer will much enhance the AOD retrieval from TEMPO. It could be studied.