Research On Species Distribution Prediction And Computational Framework Integrating Multi-source Dat

A mine is not only a business but also a geographic entity.The ultimate goal of mine informatization is boosting both business competitiveness and environmental synergy.However,there lacks quality biodiversity data to improve environmental synergy,particularly,biodiversity protection.Informatization is not combined with Landscape Ecology,Biogeography,and Geoinformatics.Consequently the high spatial resolution,quantified,and spatially continuous species distribution data are not available,resulting in the lack of required data,technique and tools to realize the idea of Green Mining.This article invented a multi-source data based species distribution model,designed the integrated computing framework for multiple species distribution forecast,and applied both in a case study of Yunnan Province,China,which is rich in both biodiversity and mine resources,resolving the availability issue of quality biodiversity data and supporting tools and enabling informtization to contribute more in the green development of mine industry.The following researches were performed.(1)The acquisition of species distribution scopes.A species distribution model was invented based on the integrated multi-source data,improving the forecast performance through the combination of various species distribution data.To verify its feasibility and reliability,the model was used to study the distribution scopes of Rhododendron wardii,in comparison to other traditional models.(2)The integrated computing framework for species distribution forecast.Based on GIS,R language,and application component programming technique,the framework integrated species data,species distribution data,environment data,and prediction algorithm,meeting the data processing and analyzing requirements and automating the data preparation and model computation.On the top of single species distribution forecast,it realized the general spatial analysis such as overlays of multiple species distribution forecast outcomes,hot spot identification,and protection absence analysis.(3)A method that quantifies the impact of mine resource development on the biodiversity.It was species distribution data based,species oriented,and implemented at the levels of single species distribution and species richness.(4)A case study of the impact of mine resource development on plants in Yunnan Province,China.The species distribution data and environmental data were collected for plants in Yunnan.The spatial distribution scopes were forecast for 8654 botanical species.The overlays at 1 km spatial resolution were generated for botanical species,endemic species and threatened species,with their richness hop spots identified,protection absence analyzed.The spatial relationship between the mine resource development and botanical diversity was analyzed and the impact was calculated.The above researches yielded the following outcomes.(1)Multi-source data based species distribution forecast model MSDM,which extends traditional species distribution models and includes innovative methods such as the algorithm of pseudo absence spots based on environment and spatial distance,priori distribution based boolean distribution threshold search strategy,and priori distribution based forecast accuracy evaluation method.Those methods combines the advantages of various species distribution data,resolved the problems of traditional species distribution modeling regarding sampling bias,uncertain threshold,and performance evaluation,and provides a feasible approach to acquire the huge amount of species distribution data in and around mines.(2)A computing framework for multiple species oriented distribution forecast model,which is featured with streamlined data processing,the visualization and spatial analysis of forest outcomes,capability in handling massive species,and improved computing efficiency.It provides the scientific evidences and feasible methods in applying species distribution model theories and methods to support the decisions on biodiversity protection.(3)A species distribution data based impact evaluation method,EIDSDD.It utilizes species distribution data and evaluates the impact of mine resource development on plants through the scope,degree and time three dimensions of the impact on the distribution and richness of each species.The method quantifies the impact and is easy to use.The outcomes can be used to measure the different impact on plants by various development plans or development projects.(4)The botanical richness layout of Yunnan Province was presented in a spatially continuous manner at a species level,with the impact of mine resource development quantified.3.46% of mining area and 3.15% of prospecting area were in regions of 7 or higher level of botanical richness.5.56%,3.20%,and 2.75% of mining area and 4.39%,1.89%,and 1.54% of prospecting area were in the hot spots of overall botanical species,endemic species,and threatened species,revealing the spatial conflicts of mine resource development and plant protection.The nature reserves covered only a small part of hot spots,27.41% for overall botanical species,28.23%for endemic species,and 29.40% for threatened species.In their 10 km buffer regions,there were 23.09%,11.93%,and 10.28% of mining area respectively.The findings indicated there was obvious absence of plant protection and sole avoiding mining in nature reserves was insufficient to reduce the negative impact of mine resource development.The impact indices for overall botanical species,endemic species,and threatened species were 68.57,3.52,and 1.68 for the whole Yunnan province.From the view of region,the north east of Yunnan presented a higher impact index.From the view of time,the impact indices had increased steadily all time for all regions except the decrease of the west of Yunnan in 2019.From the view of species,some rare and endangered species faced greater survival pressure.