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. 相似文献
In recent years, many coal-producing countries have paid great attention to the land subsidence causedby coal cutting. In China, because of the dense population in coalfield areas, the land subsidence hazard is more seri-ous. After a brief analysis on the mechanism of land subsidence, this paper gives a comprehensive and systematical ac-count on all kinds of hazards caused by the land subsidence in China. The study shows that land subsidence has endan-gered land, buildings, traffic and communication lines, dykes and dams. It also causes damage to ecological and socialenvironment. In order to lessen the hazard of land subsidence, preventive measures should be taken to reduce the col-lapse amount, such as extraction with stowing, banded mining system, succession and coordination mining system, orhigh-pressure mudflow between rock strata. Measures of reinforcing or moving certain buildings should also be taken toreduce the destructive degree. In order to harness the subsidence land and bring them under control for fanning, mea-sures should be taken such as filling with spoil or fine breeze, excavating the deeper and covering the shallower land. 相似文献
Long-term measurement of carbon metabolism of old-growth forests is critical to predict their behaviors and to reduce the uncertainties of carbon accounting under changing climate. Eddy covariance technology was applied to investigate the long-term carbon exchange over a 200 year-old Chinese broad-leaved Korean pine mixed forest in the Changbai Mountains (128°28′E and 42°24′N, Jilin Province, P. R. China) since August 2002. On the data obtained with open-path eddy covariance system and CO2 profile measurement system from Jan. 2003 to Dec. 2004, this paper reports (i) annual and seasonal variation of FNEE, FGPP and RE; (ii) regulation of environmental factors on phase and amplitude of ecosystem CO2 uptake and release Corrections due to storage and friction velocity were applied to the eddy carbon flux.
LAI and soil temperature determined the seasonal and annual dynamics of FGPP and RE separately. VPD and air temperature regulated ecosystem photosynthesis at finer scales in growing seasons. Water condition at the root zone exerted a significant influence on ecosystem maintenance carbon metabolism of this forest in winter.
The forest was a net sink of atmospheric CO2 and sequestered −449 g C·m−2 during the study period; −278 and −171 gC·m−2 for 2003 and 2004 respectively. FGPP and FRE over 2003 and 2004 were −1332, −1294 g C·m−2. and 1054, 1124 g C·m−2 respectively. This study shows that old-growth forest can be a strong net carbon sink of atmospheric CO2.
There was significant seasonal and annual variation in carbon metabolism. In winter, there was weak photosynthesis while the ecosystem emitted CO2. Carbon exchanges were active in spring and fall but contributed little to carbon sequestration on an annual scale. The summer is the most significant season as far as ecosystem carbon balance is concerned. The 90 days of summer contributed 66.9, 68.9% of FGPP, and 60.4, 62.1% of RE of the entire year.
Analysis of Zn, Cu, Pb, Co, Cr, Li, Ni, K, Al, Fe extracted by 1 mol/L HCl or 0.5 mol/LHCl/H_2O_2, showed concentrations of Zn, Cu, Pb, Co, Cr, Fe, Ni were significantly correlated with Li, Al,K, and clay. Two methods are used to indicate the background value of the non-residual phase of elementsin sediments, and are the same as the methods used to indicate the background value of totalconcentrations in sediments. The first method uses correlograms and regression equations,the second usesthe mean element concentrations normalized with grain size. Li, Al, K can be used as reference elements to determine the background value of Zn, Cu, Pb, Co,Cr, Ni, Fe, while the clay concentration's correlation with some extractable concentrations can be used tocalculate the background value of the non-residual phase of elements as a percentage of clay concentrationin the sediments. Based on this study, the concept of using the background value of the non-residualphase of elements to compare the pollution level in differ 相似文献