121
An observational analysis of middle tropospheric stable layers over the tropical western Pacific ARM sites
This observational study documents the frequency and properties of mid-tropospheric stable layers using 6 years of data from the Atmospheric Radiation Measurement (ARM) Program sites located in the TWP: Darwin, Australia; Manus Island, Papua New Guinea; and Nauru Island, Republic if Nauru. We use upper air soundings to examine the prevalence of mid-tropospheric stable layers at each TWP ARM site and examine how the frequency of occurrence changes in each season and with time of day. We then employ surface observations and estimates of cloud fraction from the MMCR and MPL instruments to explore the mean properties of the troposphere associated with the presence of a stable layer. To this end, we composite cloud fraction, precipitation, and downwelling longwave radiation around the occurrence of each mid-tropospheric stable layer. We find that melting-level stable layers are common in soundings from all three sites. At Manus, stable layers are common year-round, exhibiting little seasonal variability. At Darwin, stable layers are most common during the inactive phase of the North Australian Monsoon (NAM), which is characterized by relatively little deep convection and lower cloud fractions. At Nauru, stable layers are less common during boreal spring and fall months (when the ITCZ is in the immediate vicinity) than during summer and winter months. All three sites exhibit nearly equal occurrence of stable layers in 0000 UTC vs. 1200 UTC soundings, an indication that the diurnal cycle may have little influence on the frequency of occurrence. Composite precipitation, longwave radiation and cloud fraction suggest that there is a complex relationship between the presence of deep convection and the formation and maintenance of the melting-level stable layer. In our presentation, we will explore the implications of our results for improved understanding of how mid-level stable layers affect the tropical climate system, and suggest how this information can be used to improve cloud parameterizations in global climate models.