Publication Date

2009-01-01

Availability

Open access

Degree Type

Thesis

Degree Name

Master of Science (MS)

Department

Meteorology and Physical Oceanography (Marine)

Date of Defense

2009-04-10

First Committee Member

Dr. Brian Mapes - Committee Chair

Second Committee Member

Dr. Paquita Zuidema - Committee Member

Third Committee Member

Dr. Chidong Zhang - Committee Member

Fourth Committee Member

Dr. Stefan Tulich - Outside Committee Member

Abstract

With the launch of CloudSat, direct observations of cloud vertical structure became possible on the global scale. This thesis utilizes over two years of CloudSat data to study large-scale variations of clouds. We compose a global data set of contiguous clouds (echo objects, EOs) and the individual pixels comprising each EO. For each EO many attributes are recorded. EOs are categorized according to cloud type, time of day, season, surface type, and region. From the categorization we first look at gross global climatology of clouds. Maps of cloud cover are subdivided by EO (cloud) type, and results compare well with previous CloudSat work. The seasonality of cloud cover is also examined. Focus topics studied in this thesis include: (1) mid-level clouds, (2) stratocumulus clouds, and (3) clouds across the Madden-Julian Oscillation (MJO). The mid-level cloud work found an unexpected frequency peak in EO top heights between 7-8 km in the tropics, further shown to correspond to a global peak in EO top temperature between -15°C ? -20°C. Hypotheses are discussed regarding cause of this feature. Stratocumulus clouds are defined as low-level (tops < 4.5 km), wide (width > 11 km) EOs. Stratocumulus cloud cover agrees (with understandable differences) with other estimates (ISCCP and CALIPSO). The seasonal cycle of stratocumulus over the main stratocumulus decks is examined. The Peruvian and Namibian decks have increased cloud cover in austral spring in 2007 vs. 2006, corresponding sensibly to sea surface temperature differences and changes in lower static stability. Looking at rain and drizzle statistics, wider EOs are found to drizzle more. Clouds across the MJO are defined relative to temporally filtered OLR data. Cloud cover (volume) doubles (triples) from suppressed to active MJO phases, with some shifts of the relative contributions of different EO types from the front to back of the MJO. Pixel statistics in dBZ-height space correspond to these cloud-type shifts. High anvils and low clouds in front lead deep convection followed by relatively lower anvils in the back.

Keywords

Cloud Microphysics; Ice Crystal Growth; CFAD; El Nino;

Share

COinS