基于光学厚度光谱特征改进北京地区气溶胶积聚模态比例估计

积聚模态比例(AMF)是细模态气溶胶光学厚度占总光学厚度的比例，是估算直接辐射强迫和评估细颗粒物污染的重要参数。利用2011年北京地区AERONET网站提供的气溶胶光学厚度(AOD)和粒子体积谱分布等微观参数，计算极值修正后的细模态Angstrom指数(α_f)并进行气溶胶光学厚度光谱退卷积计算，改进AMF估算精度。AMF估算误差主要是由于α_f的误差估计偏小造成的。通过统计模拟的α_f极大值与极小值范围获取约束条件，建立极值修正方法。通过敏感性试验可知，极值修正可将α_f约束在合理范围内，且α_f和AMF对气溶胶归一化体积谱分布敏感，不同谱分布可使α_f从0.662变化到2.849，AMF从0.08变化到0.84。经过极值修正后，AMF的平均误差从0.072减小到0.044，减小了38.89%，尤其在冬夏两季，误差减小明显。AMF精度的提高可直接提高细模态及人为气溶胶光学厚度的估算精度，对人为气溶胶的直接辐射强迫估算、气候变化评估和环境质量评价有着重要意义。

Improving Accumulation-Mode Fraction Based on Spectral Aerosol Optical Depth in Beijing

In the present study, the authors put forward a parameterization method of correcting fine-mode Angstrom index with aerosol optical depth and volume distribution of aerosol derived from AERONET in Beijing over 2011. The parameterization method is coupled with aerosol optical depth spectral deconvolution algorithm to improve the accuracy of accumulation-mode fraction. The errors of estimated AMF are derived from underestimate of fine-mode Angstrom index errors. We calculate and simulate the extreme values of fine-mode Angstrom index, getting constraint conditions, and then establish the extreme values correction method. Results from sensitivity test suggest that extreme values of fine-mode Angstrom index are constrained in the reasonable range. Fine-mode Angstrom index and AMF are sensitive to normalized volume distribution of aerosol, ranging from 0.662 to 2.849 and from 0.08 to 0.84 due to different distribution. Mean deviation of accumulation-mode fraction is reduced from 0.072 to 0.044, and the difference is 38.89%, especially in winter and summer. Improving computational accuracy of accumulation-mode fraction can enhance that of anthropogenic aerosol optical thickness, and it has important significance of anthropogenic aerosol direct radiation force estimation and environmental quality assessment.