免疫进化算法反演均匀混合气溶胶吸湿增长因子

通过对大气消光系数进行组分分解，并借助米散射理论，构建了以均匀混合气溶胶吸湿增长因子为唯一变量的目标函数.进一步利用免疫进化算法优化该目标函数，提出了一种针对均匀混合气溶胶吸湿增长因子的反演算法.基于成都市2017年10~12月浊度计，黑碳仪和GRIMM180环境颗粒物监测仪的地面逐时观测资料以及该时段同时次的环境气象监测数据（大气能见度，相对湿度RH和NO₂质量浓度），评估了算法的性能及其适用性.结果表明：对所有测试样本而言，反演均匀混合气溶胶吸湿增长因子的免疫进化算法均能快速收敛到全局最优解.建立了成都地区秋冬季均匀混合气溶胶吸湿增长模型，该模型显著提升了环境条件下气溶胶散射系数的模拟精度，其模拟值与实测值之间的平均相对误差仅为12.7%.该反演算法的普适性可为气溶胶吸湿性及其辐射强迫效应的后续研究提供算法保障.

Retrieval of hygroscopic growth factorof uniformly mixedaerosol particles based on immune evolution algorithm

Based ondecomposition of atmospheric extinction coefficient and Mie scattering theory, the objective function, which has the solevariable of thehygroscopic growth factor of uniformly mixed aerosol particles, was established.Furthermore, immune evolution algorithmwas further used to optimize the objective function, and afeasiblemethod was proposed to retrieve hygroscopic growth factor of uniformly mixed aerosol particles. The performance and applicability of the method was evaluatedby utilizing the hourly ground observation data fromnephelometer, aethalometer and GRIMM180 environment particle monitorin Chengdu from October 2017 to December 2017, as well as the coincidental environmental and meteorological data, whichincludesatmospheric visibility, relative humidity (RH) and NO₂ mass concentration. The results suggest that the retrieval algorithm is characterized by fast convergence, robustness and precision for all tested samples. Hygroscopic growth model of uniformly mixed aerosol particles was established during autumn and winter in Chengdu. Note that this modelcould significantly improve the simulation accuracy of aerosol scattering coefficient in ambient conditions. The advantage of this model is that the average relative error between the simulated and the measured is only 12.7%. The universal algorithm isbeneficial for subsequent study on the aerosol hygroscopic properties and its radiative forcing impacts.