Aerosol and Cloud Relationships Revealed by the Multi-angle Imaging SpectroRadiometer (MISR) Instrument

After decades of study, the effects of aerosols on clouds remain the largest source of uncertainty for future projections of climate change. The past few years have seen extensive use of satellite data to attempt to address this challenge with global aerosol and cloud retrievals. These studies typically make use of Level 3 gridded data at 1° x 1° resolution, often from the Moderate Resolution Imaging Spectroradiometer (MODIS) instrument, which flies on NASA’s Terra and Aqua satellites. Here we consider the complementary information on aerosols and clouds provided by the Multi-angle Imaging SpectroRadiometer (MISR) instrument that accompanies MODIS on Terra. MISR’s nine cameras observe in the visible and near infrared, providing information on aerosols over all types of surfaces and allowing the height of clouds to be derived using geometric techniques, which results in high vertical and spatial resolution. MISR cloud top height retrievals are particularly sensitive to low-level cumulus and stratocumulus clouds, where other satellite retrieval techniques can have difficulties due to the thermal structure of the atmosphere, and MISR retrievals are less sensitive to thin cirrus that can obscure lower-level cloud fields more likely to be influenced by aerosols.

In this study we consider MISR Level 2 (swath-based) aerosol and cloud datasets at 17.6 km and 1.1 km spatial resolution, respectively. This information can be aggregated into Level 3 (globally gridded) products on a daily, monthly, seasonal, or yearly basis; but this aggregation can have important consequences on the conclusions derived from these datasets. First, simple averaging discards potentially significant information regarding the statistical distribution of the geophysical variable of interest. Second, relationships between two or more variables are likely to be further distorted by the aggregation and averaging process, particularly when the variables are derived independently at different initial resolutions. This is demonstrated by comparing average cloud top height as a function of mean aerosol optical depth – a relationship that may be indicative of convective invigoration of clouds by aerosols. The details of the behavior of satellite datasets paired in various ways can be significantly different, which has important implications for studies using daily or monthly Level 3 data to examine aerosol effects on clouds. We will discuss which relationships best represent the likely true connection between these and other aerosol and cloud geophysical parameters in the climate system.