The Dependence of Cloud Property Trend Detection on Satellite Instrument Accuracy

How clouds will change and interact with Earth’s changing climate is one of the largest sources of uncertainty among Global Climate Model (GCM) projections. Modeling experiments from the fifth phase of the Climate Model Intercomparison Project (CMIP5) generally agree that the net cloud feedback is likely positive, indicating that as clouds change they will cool the planet less as global mean temperatures increase. However, CMIP5 models exhibit large disagreement regarding the magnitude of the net cloud feedback. The largest contributor to the inter-model spread in net cloud feedback and Equilibrium Climate Sensitivity (ECS) among CMIP5 models is the uncertainty in shortwave cloud feedback.  Reducing uncertainty in ECS in part requires a more refined understanding of the mechanisms driving shortwave cloud feedback using information from observations and climate model output. Cloud feedback can be determined from satellite observations by monitoring decadal trends in shortwave and longwave cloud radiative effect (previously, cloud radiative forcing), which is the difference between the clear and all sky top of atmosphere irradiance. An understanding of the physical mechanisms driving shortwave and longwave decadal cloud radiative effect trends is dependent upon a comprehensive understanding of how global cloud properties that govern changes in cloud radiative effect respond to changes in Earth’s climate.

Detecting trends in essential climate variables, such as the cloud properties evaluated in these studies, on the global, decadal scales relevant to climate change studies requires highly accurate and stable measurements and retrieval algorithms. Climate variable trend accuracy is dependent upon the magnitude of natural variability, instrument and retrieval algorithm accuracy and stability, and the magnitude of the trend in question. We have applied a climate change accuracy framework to quantify the impact of absolute instrument calibration on cloud property trend uncertainties. The cloud properties used in these studies are cloud fraction, effective temperature, optical thickness, and effective radius retrieved using the Clouds and the Earth’s Radiant Energy System (CERES) Cloud Property Retrieval System, which retrieves cloud properties using Moderate-resolution Imaging Spectroradiometer (MODIS) reflectances and brightness temperatures.

The climate accuracy framework also allows for determining the time to detect a trend of some magnitude using instruments with various absolute calibration accuracies. Additionally, using the forcing-feedback equation and radiative kernels, we estimated a relationship between cloud property trend accuracies, SW and LW cloud feedbacks, and ECS. We also estimated the connection between water cloud effective radius trend uncertainties and CMIP5 trends in Effective Radiative Forcing due to aerosol-cloud interactions (or, the aerosol indirect effect). This direct relationship between instrument accuracy requirements and climate model projections allows for determining the level of instrument absolute accuracy needed to reduce climate model projection uncertainty. 

Different cloud types have varied radiative impacts on the climate system depending on several attributes, such as their thermodynamic phase, altitude, and optical thickness. Therefore these analyses have also been conducted for different cloud types to provide a clearer understanding of the instrument accuracy requirements needed to detect changes in their cloud properties. This information combined with existing knowledge of the radiative impact of different cloud types can be applied to prioritize among requirements when designing future satellite sensors and understanding the detection capabilities of existing sensors.