基于大数据的西藏荒漠化治理植物优选与验证

植物种植作为荒漠化治理的重要方式之一,关系到荒漠化区域的高效可持续发展。为提高植物选择的科学性与合理性,在项目前期已建成荒漠生态治理大数据平台并实现中国主要荒漠化区域生态信息全方位采集的基础上,进行基于大数据的植物优选研究与试验验证。研究运用相关性分析、聚类分析等大数据分析方法对荒漠植物种质资源库中植物进行类别划分,初步筛选适应该地区气象条件的植物;进一步运用层次分析法、专家打分法等决策方法从土壤、地形、生态效益、经济效益和其他等5个方面进行综合分析与评价,以优化初选结果。将该方法应用于西藏地区荒漠化治理植物的选择,经大数据挖掘分析,初步筛选出了核桃、黑果枸杞、盐生草和花花柴等适宜植物;进一步优化分析得出,核桃具有经济效益高、耐储运、前期投入相对较少等优势,是该区域荒漠化治理中生态适应性与综合效益俱佳的植物。优选结果在西藏山南市扎囊县桑耶镇的荒漠化治理中得到了验证,目前核桃长势良好,预期生态与经济效益显著。基于大数据进行荒漠治理植物的优选可为荒漠化区域科学规划及高效治理提供坚实的理论与数据支撑。

Optimal selection and verification of plant species for desertification control in Tibet based on big data

Land desertification has posed a major hazard to the society, economy, and environment in the Tibet semi-arid areas. The scientific and rational control of desertification becomes much more important to improve the living environment of human beings and the sustainable development of ecological system. Plant planting is expected to be one of the effective means of desertification governance. Previous research has conducted on the selection of plants species for the desert governance, but it is necessary to accurately optimize the specific plant species, particularly on the range of plant selection, data analysis process and consideration of comprehensive benefits. In this paper, a method of plant selection was proposed based on big data. Two steps were mainly included: one step is to optimize plants that meet the climate suitability of the study area in preliminary analysis, and another step is to verify the influence factors of the selected plants in comprehensive evaluation. A platform of big data was established for desertification ecological governance and the realization of all-round collection of major desert ecological information in China. The initial conditions were then defined according to the characteristics and direction of desertification governance in the study area. A database of germplasm resources for desert restoration plant was constructed to classify and select the subsequent plant categories. It is also necessary to consider the limitations of the dominant meteorological factors, such as "light, temperature, and water", due to the relatively serious problems of water shortage, high intensity of sunshine, and low accumulated temperature in the study area. After the data was automatically collected by the Internet of Things (Iots), the meteorological characteristics and changes of the study area were used to provide theoretical support for the further matching of plant varieties with high temperature resistance, high intensity of sunlight and strong drought resistance. Plants with similar properties were classified into a group based on the threshold value of suitable meteorological conditions of plants by using cluster analysis, correlation analysis and other big data methods. The adaptability of plants was also analyzed during this time. In clustering, the upper threshold was 4, while the lower threshold was 5. The correlation coefficient was calculated between the plants'' own suitable environment and the meteorological conditions in the study area, and the average value was obtained by category. In the plant categories with a correlation coefficient greater than 0.95, the plants that located in both upper and lower classifications were assumed as optimum match on the meteorological conditions in the study area. The second step was to select plants with a high degree of comprehensive suitability based on the preliminary generation scheme. Taking the preliminary selected plants as the evaluation object, an expert scoring method and analytic hierarchy process were used to make a horizontal comparison and comprehensive ranking of plants, particularly on considering the influence of topography, soil, ecological benefits, economic benefits, farmers'' planting preference and policy support on the growth of plants. The experimental results showed that Juglans regia L. scored the highest, Lycium ruthenicum Murr. was the second, while Halogeton glomeratus and Karelinia caspica scored were the lowest. Since four plants have good wind and sand fixation, and soil improvement effects, Juglans regia L. planting has the advantages of less initial investment, storage resistance, and easy long-distance transportation, indicating that it was suitable for the economic and ecological benefits for the study area. Therefore, the introduction of Juglans regia L. can be the optimal option for desertification governance in the study area. The results of plant selection were also verified in the experimental areas of Sangye Town, Zhanang County, Shannan City, Tibet. In the field county, the desertification phenomenon has been severe for a long time, with the vegetation types of mainly herbaceous plants, where the survival period of plant is relatively short, even no living trees in some areas. The flat areas surrounded by mountains on three sides can be set as the site, where can be more suitable for the growth of early Juglans regia L.. Therefore, Juglans regia L. has been planted for more than two years under the intervention and management of artificial science. At present, good growth and strong branching ability show an optimal ecological adaptability to the natural environment of the area. Compared with other plants grown in the same period, Juglans regia L. has better performance on the resistance of high temperature, sunshine and drought. In addition, the number of herbs around walnuts has increased significantly in more than two years, to achieve the expected ecological and economic benefits. Both theoretical analysis and actual trial experiments have demonstrated that it is promising feasible to plant Juglans regia L. in the areas of Sangye Town, Zhanang County, Shannan City, Tibet.