Search

An advanced software library supporting parallel unstructured adaptive mesh refinement for large-scale scientific and engineering simulations.

We will explore our experiences and evaluate the performance impact this hardware introduces for our system.

The impact of gigabit networking with Myrinet 2000 hardware and MPICH-GM software on a 2-way SMP Beowulf-class cluster for parallel unstructured adaptive mesh refinement using the PYRAMID library is described.

An approach for modernizing legacy scientific software, based on the new features of Fortran 90/95, is presented.

In this paper we outline the results of using three generations of cluster machines at JPL for a select set of applications drawn from a larger development effort.

UNKNOWN

Adaptive mesh refinement (AMR) calculations carried out on structured meshes play an exceedingly important role in several areas of science and engineering. A strategy for using Fortran 90 in an object-oriented fashion is presented. This permits AMR applications to be expressed in terms of familiar abstractions that are natural to the process of solving AMR hierarchies. The OpenMP features that are useful for parallel processing of AMR hierarchies in a load balanced fashion on multiprocessors is described.

Legacy software has great value since it is generally well debugged, produces results that are trusted and is actively meeting end-user goals. The amount of hidden expert knowledge embedded in such software can be significant, making its preservation important.

UNKNOWN

The design and implementation of Pyramid (http://www-hpc.jpl.nasa.gov/APPS/AMR/), a software library for performing parallel adaptive mesh refinement (PAMR) on unstructured meshes, is described.

One goal of the Numerical Turbulent Transport Project is to model a tokamak (fusion) plasma with 10(sup 8 - 10)(sup 9) particles, to explain anomalous transport of particles and energy.

Fortran90 is a modern, powerful language with features that support important new programming concepts, including those used in object-oriented programming.

r summarizes techniques for emulating in Fortran90 the most important object-oriented concepts of C++ classes (including abstract data types, encapsulation and function overloading), inheritance and dynamic dispatching.

A final progress report is presented on the Centralized Storm Information System (CSIS). The primary purpose of the CSIS is to demonstrate and evaluate real time interactive computerized data collection, interpretation and display techniques as applied to severe weather forecasting. CSIS objectives pertaining to improved severe storm forecasting and warning systems are outlined. The positive impact that CSIS has had on the National Severe Storms Forecast Center (NSSFC) is discussed. The benefits of interactive processing systems on the forecasting ability of the NSSFC are described.

Data collected by the lightning data concentrator are available for research. The Mark 3 McIDAS capability provides greater flexibility for the Marshall user community and serves as a model of future UW McIDAS to remote computer links. Techniques were investigated for the display of dynamic 3-D data sets. To date the most promising display technology is a polarized two CRT perspective display which allows both dynamic 3-D images and graphics presentations with full color capability. Algorithms were for the preparation and display of conventional and satellite based weather data in 3-D. These include gridding, contouring, and streamlining processors which operate on both real time and case study data bases. An upper air trajectory model was implemented which creates a display of air parcel trajectories in perspective 3-D. A subsystem for the generation of 3-D solid surface display with shading and hidden surface display with shading and hidden surface removal was tested and its products are currently being evaluated. Motion parallax introduced by moving the point of observation during display is an important depth cue, which, when added to the perspective parallax creates a very realistic appearing display.

Applications Technology Satellite (ATS) 3 green sensor data are used to measure surface reflectance variations in the Sahara/Sahel during the recent drought period 1967-74. The magnitude of the seasonal reflectance change is shown to be as much as 80% for years of normal precipitation and less than 50% for drought years. Year-to-year comparisons during both wet and dry seasons reveal the existence of a surface reflectance cycle coincident with the drought intensity. The relationship between the green reflectance and solar albedo is examined and estimated to be about 0.6 times the reflectance change observed by the green channel.

Applications Technology Satellite 3 green sensor data were used to measure surface reflectance variations in the Sahara/Sahel during the recent drought period; 1967 to 1974. The magnitude of the seasonal reflectance change is shown to be as much as 80% for years of normal precipitation and less than 50% for drought years. Year to year comparisons during both wet and dry seasons reveal the existence of a surface reflectance cycle coincident with the drought intensity. The relationship between the green reflectance and solar albedo is examined and estimated to be about 0.6 times the reflectance change observed by the green channel.

Applications Technology Satellite 3 data are used to measure surface albedoes in the African Sahel during the 1967 to 1974 drought.