Evaluation Of The 2030 Sustainable Development Goals From The Perspective Of Geographic Space

The UN 2030 Agenda for Sustainable Development is a major global public policy and political program that guides the development of countries in the world over the next decade.The agenda proposes 17 Sustainable Development Goals(SDGs)and 169 targets covering the three major areas of economy,environment and society.Quantitative assessment and monitoring of the implementation of SDGs is an important measure to implement the 2030 Agenda for Sustainable Development,and is also a frontier and hotspot in academic research both at home and abroad.The evaluation and monitoring of SDGs involve many factors and complicated technologies,and are still in the stage of conceptual design,method discussion,and singleindicator,small-scale pilot.The main technical problem is the quantitative calculation of indicators integrating geographic information and statistical data and the quantitative assessment of regional SDGs under the framework of global unified indicators.This article takes SDG6(Ensure availability and sustainable management of water and sanitation for all)as an example,in a complete administrative area(Deqing County),to explore the monitoring and evaluation methods of regional SDGs,based on the establishment of a localized indicators system.In the above,using Earth Observation Technology(EO)and Geographic Information System(GIS),we focused on the spatial monitoring methods of typical indicators to achieve the quantitative evaluation and comprehensive evaluation and analysis of all localized indicators of SDG6.This article provides reference methods and examples for the comprehensive evaluation and related research of SDG6 carried out in other regions at home and abroad.The main research results are as follows:(1)Combining China's country plans and the actuality of the study area,with the principles of adaptability,measurability,and full coverage,this paper proposed the SDG6 localization indicator system(including 7 indicators and 11 indicators)based on the UN SDGs Global Indicator Framework(SGIF)and developed three spatial monitoring methods to measure key indicators that are difficult to evaluate through traditional statistical data,and combined with statistical methods to achieve the quantitative calculation of all SDG6 indicators.At the same time,this paper also proposes a comprehensive evaluation method of indicators and goals at the county scale.Practice has proved that the regional SDG6 assessment and monitoring method proposed in this paper is feasible.(2)Aiming at the problem of efficient and accurate extraction of water resources information in SDG6 medium and long time series,this paper proposes a method for automatic extraction of water body information based on time series sample library.Based on long-term Landsat satellite data,considering different types of water bodies(primary water system,secondary water system,lake,lake center,lake edge,junction of river and lake,etc.),different meteorological conditions(no cloud,thin cloud,thick cloud),established Long time series water sample library(including annual season,rainy season,dry season,quarterly,monthly,thin cloud,etc.).According to the characteristics and needs of different regions,the sample library is automatically matched according to certain rules(such as satellite imaging time),and machine learning algorithms(such as random forest)are used to achieve fast and automatic extraction of long-term water resources information.(3)Aiming at the problem of long-term surface coverage information extraction in waterrelated ecosystems,this paper proposes an automatic extraction method of long-term surface coverage of the spectral features of period features.Based on the long-period periodic characteristics of a single pixel,through the sampling and fitting calculation of stable samples of different surface coverage types,a sample set of fitting parameters of different land types is formed,and the method of random forest is used to establish models for fitting parameters of different features.The CCDC algorithm is used for continuous change detection of each pixel of the long-sequence Landsat image in the study area,to determine the mutation point(time)of the single pixel feature category,and to determine the parameters of the fitting curve before and after the mutation point.Then enter these parameters into the random forest model for classification to determine the type of features before and after the mutation point.With the help of the powerful computing function of the cloud platform Google Earth Engine(GEE),the long-term feature categories of the pixels are extracted one by one to realize the automatic extraction of long-time surface coverage types.(4)Aiming at the problem of public health facility service scope and population evaluation and calculation,this paper proposes a convenience model of public service facilities that takes into account travel modes.The spatial distribution of urban public toilets in the study area is obtained through the spatial data of municipal toilets and combined with the source data(Baidu,Gaode,etc.).The accessibility of public health facilities is calculated by the urban road network data,and considering the different travel wishes of urban population,a method for calculating the convenience of public health facilities services that takes into account the travel mode(walking or riding)is established.Finally,combined with population density data,an evaluation result with "spatial-temporal coordinates" is formed,and the index calculation result is dynamically expressed.