Mid-level clouds have been studied less than their high and low counterparts and their characteristics and radiative effects are not well understood. Existing satellite-based cloud climatologies, such as ISCCP, using passive remote sensors place less emphasis on mid-level clouds, which might cause disagreements with climate model simulations and radiative forcing of such clouds. A path forward to improve the evaluation of cloud feedbacks needs to advance the representation of clouds in the current climate state. Therefore, accurate characterizations and comparisons between observations and model output are quite important.

We will present a new characterization of mid-level clouds using combined active and passive remote sensing techniques from A-Train satellites, including CALIPSO and CloudSat, and ARM ground-based observations. First, we will test the sensitivity of mid-level cloud frequency of occurrence to different ways of defining where the mid-level is. Specifically, we will see how perturbations to the traditional definitions of the mid-level as occurring between cloud top pressures of 680 and 440 hpa or cloud top heights between 3 and 7 km affect mid-level (and high and low) cloud frequencies. These tests will be performed both in the Tropical Western Pacific region and globally. In addition, the usefulness of cloud top temperature as a midlevel cloud definition will be investigated. The new characterization including cloud boundaries, cloud top pressure and temperature will be demonstrated for single- and multi-layered cloudy scenarios separately. The likelihood of the occurrence of mid-level clouds in multi-layered cloud systems, which can result in serious biases in characterizing mid-level clouds, will also be evaluated.