Breakout Summary Report

ARM/ASR User and PI Meeting

2 - 6 May 2016

Session Title: Advances in aerosol and cloud retrievals from spectral radiometers
Session Date: 2 May 2016
Session Time: 3:30 PM - 5:30 PM
Number of Attendees: 0
Summary Authors: Laura Riihimaki, Connor Flynn

Breakout Description

Saved for now.

Main Discussion

The session was introduced as a time to update the broader community on key progress related to ARM spectral radiometers and derived retrievals, and a chance to discuss future priorities for hyper-spectral radiometers. The session was well attended with interaction on all topics. Much of the discussion focused on scientific needs and data quality of hyper-spectral instruments. To continue discussion beyond the meeting, Connor will develop a table describing/characterizing the relative importance of different corrections to measurement data quality, and Laura will create a spectral measurements email list. The session agenda is included here:
Updates on Spectral Radiometers and Retrievals
• Laura Riihimaki—MFRSR 1625 nm channel
• Rick Wagener—Implementing Christine Chiu’s 3-channel Cimel cloud retrieval
• Justin Monroe—New VAP data-quality efforts supporting MFRSR, spectral shortwave
• Dave Turner—New AERIoe retrieval
• Sasha Marshak—Spectral cloud retrievals from MAGIC
• Joe Michalsky—RSS data/aerosol retrievals from Aug 2009-Mar 2014.

Hyper-spectral Shortwave Retrievals and Data quality
• Connor Flynn—update on SASHE/SASZE data quality and current operational AOD retrievals
• Discussion about priorities/path forward for SAS data quality and other spectral radiometer needs.

Key Findings

Key findings related to hyper-spectral shortwave measurements:
• Results from MAGIC study using SSFR, SASZE, and CIMEL (Marshak et al.) showed differences in the magnitude of the values from the instruments could be large (~10-30%), but spectral dependence of the differences is weak, allowing for a clever retrieval using the slopes and intercepts of spectral relationships. Analyzing the SASZE and SSFR measurements of the cloud/clear transition zone during MAGIC, Marshak et al. found that inhomogeneous mixing dominates (no substantial changes in cloud drop size) near cloud edges.
• The UV spectral range has a strong absorption signal, so can give new insight into aerosol absorption. Michalsky demonstrated that one can obtain 2-sigma uncertainties in retrieved single-scattering albedo of +/- 0.03 down to AODs of at least 0.08 (i.e., much lower than the CIMEL 0.40 limit) with a UVRSS in the 300-360 nm region when using the instrument in transmission mode, which avoids the need for absolute calibration.
• Examination of SASHE AOD retrievals at SGP shows cases with steady AOD retrievals during the day, indicating good spectral response. Other instrument comparisons over time have also shown good results.
• Justin Monroe from the data quality office showed results comparing AOD retrieved from SASHE to that from MFRSR. Initial results show alignment errors and changes in AOD quality over time, which are starting to be examined. After the session, he received a request from an ASR PI for a similar comparison at another site and was able to show good quality of AOD from the SASHE at that site.

Needs

• There is a clear need for additional comparisons between different shortwave spectral instruments to better assess the current data quality. Some planned activities were discussed, but additional comparisons may be needed. Planned activities include:
o Compare zenith radiances from SASZE and CIMEL cloud mode radiances, now possible due to new ingest of CIMEL cloud mode radiances from AERONET; Justin Monroe may expand DQO plots to include these additional measurements as they become available.
o Comparison between RSS and SASHE for period of time when both were working.
• Much of the discussion focused around what hyper-spectral measurement accuracy is needed for given science applications and what science can be done with current accuracy. Some uncertainty remained at the conclusion of the session about the importance of different corrections, though some of the discussion can be summarized by what science applications can be done with accurate spectral “shape” versus absolute calibrations:
o For zenith radiances: evidence of reliable spectral shape in measurements (from comparisons between multiple instruments) can be used for clever cloud retrieval techniques like that shown by Marshak et al. It is likely the SASZE data is already sufficient for this purpose, as seen during MAGIC, though further comparisons with other radiances will improve confidence. Ongoing monitoring of SASZE by comparison with CIMEL radiances, etc. is needed to ensure that this is true at other sites and times. There is still a need for better absolute accuracy of these measurements for applications like multi-instrument retrievals (interest from Chiu, Schwartz, and likely others). For example, Christine Chiu needs accuracy of ~5% to include the data in her ENCORE retrievals. Achieving this level of accuracy with the SASZE is still a challenge and would require improved calibration techniques through either more careful lamp calibrations or a new idea pinning irradiance and radiance calibrations via the direct irradiance component and spectralon blanks.
o For hemispheric irradiances: In this case, reliable transmittance data from Langley calibrations can be used for aerosol (and some cloud) retrievals from hemispheric irradiance (interest from Kassianov, McComiskey, Michalsky, Flynn, etc), when consistency between diffuse and direct transmittances is considered robust. The achievable accuracy of transmittance measurements is estimated to be around 1-2% when an instrument is functioning well—Michalsky demonstrated this is achievable for many wavelengths with the RSS. Absolute calibration (in W/m^2) of hemispheric irradiance is likely possible to less than 5% by scaling transmittance measurements to the TOA solar spectrum, as shown by Michalsky with the RSS. This is likely more accurate than the best we can do with lamp calibrations. This may be the most useful measurement for separating the radiative forcing of different components of the atmosphere, which is needed in units of energy (W/m^2). This level of accuracy needs to be demonstrated with the SASHE as well.
o Definition of requirements and possibilities for other applications like trace gas retrievals (interest from Feldman), other multi-instrument retrievals, and better separating the radiative forcing of different components of the atmosphere (interest from Chiu, Lubin) will require further discussion.