Developing an aerosol climate data record from MODIS, VIIRS and beyond

To identify changes in direct aerosol radiative forcing (DARF), we need global observations of aerosol properties, including aerosol optical depth (AOD). A climate data record (CDR) is a time series of measurements of sufficient length, consistency, and continuity to determine climate variability and change. To reduce DARF uncertainties, we require an AOD CDR with uncertainties of ±0.02 or better (~10% of global mean), and which can discern trends of ±0.01/decade over time series of 20-30 years. To achieve this CDR, we require a stable instrument, or series of instruments, along with the same aerosol retrieval algorithm applied to all data. The two Moderate Resolution Imaging Spectrometer (MODIS) have been flying on NASA's Terra and Aqua satellites for over a dozen years. During this time, we have fine-tuned the aerosol retrieval algorithms and data processing protocols, resulting in a well-characterized AOD product. We have worked to adjust instrument calibration and retrieval protocol, so that MODIS provides a stable aerosol environmental data record (EDR). However, these instruments are degrading, so we will have a 15-18 year record, at best. With the 2011 launch of the Visible and Infrared Imaging Radiometer Suite (VIIRS) aboard Suomi-NPP, we began a new time series. VIIRS is very similar to MODIS, in regards to its resolution, orbit and wavelength bands. While routine VIIRS aerosol products have demonstrated sufficient performance characteristics to be considered as a global EDR, these routine VIIRS products show many differences on the regional scale. Most of these differences are attributable to using a different retrieval algorithm. Can we stitch the VIIRS record onto MODIS, such that when combined will meet the requirements of an aerosol CDR? Our team has created a VIIRS retrieval that is structured like that of MODIS, and its products “look” more like those of MODIS, even at the regional scale. In this paper, we discuss the status of our consistent aerosol retrieval from both MODIS and VIIRS, and show global and regional statistics obtained from each dataset. We explain some of the remaining obstacles for merging the two datasets, and consider some metrics to describe their continuity.