Meso-scale mapping of terrestrial carbon dioxide source/sink intensity using inverse analysis of atmospheric concentration change by GA

利用遗传算法逆分析大气浓度变化绘制陆地二氧化碳源/汇强度的中尺度绘图

• 批准号: 15380180
• 负责人: MACHIMURA Takashi
• 金额: $ 5.82万
• 依托单位: Osaka University
• 依托单位国家: 日本
• 项目类别: Grant-in-Aid for Scientific Research (B)
• 财政年份: 2003
• 资助国家: 日本
• 起止时间: 2003 至 2005
• 项目状态: 已结题
• 来源: https://kaken.nii.ac.jp/grant/KAKENHI-PROJECT-15380180/
• 关键词: terrestrial CO_2 source; sink; inverse model; meso-scale; genetic algorithm; MM5

This study aimed to develop a new technique mapping terrestrial carbon dioxide source/sink intensity using an inverse model. The technique consists of the continuous atmospheric carbon dioxide concentration at one observatory, the wind field and gas diffusion simulation using a meso-scale meteorological model and the inverse analysis of source/sink intensity distribution using the genetic algorithm (GA). An atmosphere sampling system including an IRGA and an automated calibration system was developed and carbon dioxide concentration was measured continuously between December 2004 and January 2006 at Suita, Osaka. Before applying the new technique to real problems, a numerical simulation of the virtual source/sink distribution was examined in order to validate performance of the inverse analysis using GA. 4 or 16 fixed source/sinks were placed randomly in a target area of 126 km x 126 km and atmospheric concentration at a observatory was calculated using the meso-scale meteorological mo … More del (MM5) and the meso-scale objective analysis data. The inverse model using GA was capable to estimate the given source/sink distribution from calculated time series of the atmospheric concentration excepting a few parts of which source/sink had little influence on atmospheric concentration at the observatory. This suggested that the source/sink intensity of land use categories but its spatial distribution is more adequate as the problem to be solved. Therefore, the source sink intensity of each land use was estimated in the real problems. Carbon dioxide source intensity was the strongest in urban area among all land use and the vegetated land showed relatively small source both in winter and summer, however the absolute intensities were not accurate. The possible reasons were error in the atmospheric concentration measurement and the inadequate background concentration value. Relative source/sink intensity of the land uses and its seasonal change could be estimated reasonably by the new technique. Less

本研究旨在开发一种利用逆模型绘制陆地二氧化碳源/汇强度的新技术。该技术包括一个观测站的连续大气二氧化碳浓度，使用中尺度气象模式模拟风场和气体扩散，以及使用遗传算法（GA）对源/汇强度分布进行逆分析。在2004年12月至2006年1月期间，研究人员在大阪水田开发了一个大气采样系统，包括一个IRGA和一个自动校准系统，并连续测量了二氧化碳浓度。在将新技术应用于实际问题之前，为了验证遗传算法逆分析的性能，对虚拟源/汇分布进行了数值模拟。在126 km × 126 km的目标区内随机放置4或16个固定源/汇，利用中尺度气象模式（MM5）和中尺度客观分析资料计算了某天文台的大气浓度。利用遗传算法的反演模型能够根据计算的大气浓度时间序列估计给定的源/汇分布，但有少数部分源/汇对天文台大气浓度影响不大。这表明，土地利用类型的源汇强度及其空间分布更为充分，是有待解决的问题。因此，在实际问题中对各土地利用源汇强度进行了估算。在所有土地利用中，城市地区的二氧化碳源强度最强，植被地在冬季和夏季都表现出相对较小的源强度，但绝对强度并不准确。可能的原因是大气浓度测量误差和本底浓度值不足。利用该方法可以合理地估算土地利用的相对源汇强度及其季节变化。少