A machine learning approach for predicting computational intensity and domain decomposition in parallel geoprocessing

ABSTRACT

High performance computing is required for fast geoprocessing of geospatial big data. Using spatial domains to represent computational intensity (CIT) and domain decomposition for parallelism are prominent strategies when designing parallel geoprocessing applications. Traditional domain decomposition is limited in evaluating the computational intensity, which often results in load imbalance and poor parallel performance. From the data science perspective, machine learning from Artificial Intelligence (AI) shows promise for better CIT evaluation. This paper proposes a machine learning approach for predicting computational intensity, followed by an optimized domain decomposition, which divides the spatial domain into balanced subdivisions based on the predicted CIT to achieve better parallel performance. The approach provides a reference framework on how various machine learning methods including feature selection and model training can be used in predicting computational intensity and optimizing parallel geoprocessing against different cases. Some comparative experiments between the approach and traditional methods were performed using the two cases, DEM generation from point clouds and spatial intersection on vector data. The results not only demonstrate the advantage of the approach, but also provide hints on how traditional GIS computation can be improved by the AI machine learning.     

Pulsar magnetic field for an infinitely conductive accretion disk

The magnetic-field distribution outside a flat, infinitely conductive unbounded disk in the field of a point magnetic dipole is determined. A relationship is established between the problem of magnetic-field determination and the problem of the flow of an ideal incompressible fluid around an infinitely thin disk.     

Stochastic modelling of groundwater flow and solute transport in aquifers

This paper presents an introductory overview of recently developed stochastic theories for tackling spatial variability problems in predicting groundwater flow and solute transport. Advantages and limitations of the theories are discussed. Lastly, strategies based on the stochastic approaches to predict solute transport in aquifers are recommended.     

Pulsar Wind Nebulae in Egret Error Boxes

A remarkable number of pulsar wind nebulae (PWN) are coincident with EGRET γ-ray sources. X-ray and radio imaging studies of unidentified EGRET sources have resulted in the discovery of at least six new pulsar wind nebulae (PWN). Stationary PWN (SPWN) appear to be associated with steady EGRET sources with hard spectra, typical for γ-ray pulsars. Their toroidal morphologies can help determine the geometry of the pulsar which is useful for constraining models of pulsed γ-ray emission. Rapidly moving PWN (RPWN) with more cometary morphologies seem to be associated with variable EGRET sources in regions where the ambient medium is dense compared to what is typical for the ISM.     

Coronal change at the south-west limb observed by Yohkoh on 9 November 1991, and the subsequent interplanetary shock at Pioneer Venus Orbiter

A spectacular change in the lower corona on the south-west limb has been found in solar images taken by the Yohkoh soft X-ray telescope. The event is characterized by a large topological change in magnetic field and a large intensity decrease observed after the X1. 1/1B flare on 9 November, 1991. A coronal mass ejection (CME) was observed by the Mark III K-coronameter (MK3) at the HAO/Mauna Loa Observatory. Both the MK3 (white-light) and soft X-ray observations showed that one leg of this CME was located above the flare site. An interplanetary shock associated with this event was observed by Pioneer Venus Orbiter, and, possibly, by IMP-8.Also Cooperative Institute for Research in the Environmental Sciences (CIRES), University of Colorado, Boulder, CO 80309, U.S.A.