1996: Cloud Structures in Storm Lifecycles

Team Members: Kwajo Abeyie, Jose Alburquerque, Shermane Austin, John DaPonte, Robert Kruckeberg, Patrick Michel, Haidy Pena, Chris Petersen, Abdon Whitelocke, Dwayne Williams

Research Objectives 1

An analysis of the cloud type distribution in the lifetimes of ten tropical storms and the relationship of this distribution to storm strength and size characteristics.

Tools: NMC and ISCCP CX data: cloud top pressure (IPC), cloud optical thickness (TAU), sea level pressure (SLP); FORTRAN programs for producing histograms; Spyglass and Spyglass Plot.

Tasks: 1) Examine ten tropical storms by dividing the lifetime of the storm into early, middle and late stages; 2) Average the cloud properties in each stage; 3) Produce images of the spatial distribution of the average cloud properties for each stage; 4) Create histograms of the average cloud property distribution for each stage; 5) Create histograms of the cloud property distribution for each observation of selected storms.

Research Objectives 2

An analysis of the differences in cloud type distribution between storm and non-storm regions of the northern mid-latitudes in the GCM, including seasonal and interannual variability.

Tools: GCM (ISCCP CX/NMC for Contrast): cloud top pressure, cloud optical thickness, sea level pressure; FORTRAN programs for producing histograms; Spyglass and Spyglass Plot.

Tasks: 1) Examine several months of data from the GCM; 2) For each month examine the storms produced (including the number and length of storms); 3) Produce histograms of cloud property distribution for storm and non-storm areas; 4) Compare and contrast with Robert's results.

Research Objectives 3

Analyze the differences in cloud type distribution between storm and non-storm regions in northern mid-latitudes, including seasonal and interannual variability. This will allow us to make comparisons with the GCM and improve techniques of modeling clouds and cloud systems.

Tools: ISCCP CX/NMC, several years/months in a year, northern mid-latitudes: SLP, IPC and TAU; Spyglass and Spyglass Plot.

Tasks: Use FORTRAN software to create two-dimensional histograms of cloud top pressure vs. optical thickness. Use Spyglass / Excel to perform statistical operations on histogram data.

Research Objectives 4

The development of a web browser application that provides information on storms and presents results of the cloud type distributions in storms.

Tools: ISCCP CX/NMC data for various years/months in the tropics: sea level pressure, cloud optical thickness, cloud top pressure.

Tasks: 1) Translate from FORTRAN into JAVA the program for averaging the stages of tropical storms, 2) Develop a user interface using JAVA to enable easy use of the averaging program; 3) Integrate these two JAVA applications; 4) Develop a Web Page that incorporates the resulting product.

Research Objectives 5

A description of the changes in cloud type distribution during the lifetimes of selected mid-latitude storms and their relationships to changes in storm dynamics.

Tools: ISCCP CX/NMC data for northern mid-latitudes (January, April, July, and October of 1988, January and July of 1985-1987): SLP, IPC and TAU; FORTRAN programs for plotting storm tracks and producing histograms; Spyglass and Spyglass Plot.

Tasks: 1) Select a minimum of ten storms from each month based on the characteristics of the plots of the pressure profiles and the storm tracks; 2) Produce images of the spatial distribution of the cloud properties (TAU and IPC) for each observation in the storm track; 3) Produce histograms of the cloud type distribution for each of these observations; 4) Relate these distributions to the average distributions examined by Robert.

Research Objectives 6

1) Adaptive self-organizing segmentation model that can be utilized for analysis of cloud structure changes; 2) Presently, exploring output from a linear model and determining limitations (inc. computational feasibility); 3) Summarize output of prior self-organizing models.

Tools: ISCCP and NMC data sets. Time Period: January 1987. Area: Global. Parameters: cloud top pressure and TAU. Model experiments: defined above.

Tasks: 1) Develop analytical method for output from the linear model as indicated above. 2) refine tentative model for adaptive temp. segmentation and then encode and test with samples from both cloud data and other data sets.

Research Objectives 7

1) Classify mid latitude cloud images using First and Second Order statistical parameters to provide quantitative features for Pattern Recognition. Both Statistical and Neural Network Pattern Recognition models will be used. 2) Investigate the use of Image Sequence Analysis to compute Optical Flow Fields for the these images.

Tools: Spyglass, C Language and SAS.

Tasks: Both ISCCP-C1 and NMC data sets will used in mid latitude regions. Spyglass will be used to extract Cloud Optical Thickness and Cloud Top Pressure from the ISCCP-C1 files and to extract Sea Level Pressure from the NMC files. Programs in C language and SAS will be used to analyze the data sets.