关中城市大气污染监测数据

针对污染溯源与预测问题，基于关中城市群的气象和大气污染物监测数据，以西安市为主要分析对象，系统分析城市大气污染物时空分布特征与关联因素，并在此基础上考虑大气污染扩散的物理机理，使用LSTM建立时空耦合的PM2.5预测模型，提高污染物浓度预测精度。数据集包括大气污染物监测数据和气象数据，大气污染物数据包含关中城市群的11个主要城市（宝鸡、临汾、平凉、庆阳、商洛、天水、铜川、渭南、咸阳、运城、西安），以及核心城市西安市内9个区域监测数据；气象数据包含7个主要城市（运城、渭南、咸阳、西安、运城、平凉、庆阳）。污染物监测数据包括AQI监测值、PM2.5、PM10监测值、二氧化硫监测值、二氧化氮监测值、一氧化碳监测值、臭氧监测值、臭氧滑动8小时监测值。气象数据包括温度、露点温度、气压、风向、风速、1小时降雨量、6小时降雨量。

To address the issues of pollution source tracing and prediction, based on meteorological and atmospheric pollutant monitoring data of the Guanzhong Urban Agglomeration and taking Xi'an as the primary research object, this study systematically analyzes the spatiotemporal distribution characteristics and associated influencing factors of urban atmospheric pollutants. On this basis, by considering the physical mechanism of atmospheric pollution diffusion, an LSTM-based spatiotemporal coupled PM2.5 prediction model is established to improve the prediction accuracy of pollutant concentrations. The dataset consists of atmospheric pollutant monitoring data and meteorological data. The atmospheric pollutant monitoring data covers 11 major cities in the Guanzhong Urban Agglomeration (Baoji, Linfen, Pingliang, Qingyang, Shangluo, Tianshui, Tongchuan, Weinan, Xianyang, Yuncheng, Xi'an), as well as monitoring data from 9 districts within the core city Xi'an; the meteorological data covers 7 major cities (Yuncheng, Weinan, Xianyang, Xi'an, Yuncheng, Pingliang, Qingyang). The pollutant monitoring indicators include AQI readings, PM2.5 concentrations, PM10 concentrations, sulfur dioxide (SO₂) concentrations, nitrogen dioxide (NO₂) concentrations, carbon monoxide (CO) concentrations, ozone (O₃) concentrations, and 8-hour sliding average ozone concentrations. The meteorological monitoring indicators include temperature, dew point temperature, atmospheric pressure, wind direction, wind speed, 1-hour rainfall, and 6-hour rainfall.

搜集汇总

背景与挑战

背景概述

该数据集是关中城市群的大气污染监测数据，涵盖11个主要城市和西安市9个区域的污染物监测（如PM2.5、二氧化硫等）及7个城市的气象数据（如温度、风速等），用于分析污染物时空分布并基于LSTM模型进行PM2.5预测，支持污染溯源与预测研究。数据量约7.16MB，以xlsx和docx格式提供。