线翼截断方式对大气辐射计算的影响

THE EFFECTS OF LINE WING CUTOFF ON RADIATIVE CALCULATIONS

中文关键词:  逐线积分，线翼截断，吸收系数，大气透过率，大气冷却率。

英文关键词:LBL integration, Line-wing cutoff, Absorption coefficient, Transmittance, Cooling rate.

中文摘要:

      在大气辐射传输计算方法中，有3种基本方法，即，逐线积分方法，k-分布方法和带模式方法。其中，逐线积分方法是最精确的计算大气透过率的方法，本文根据透过率计算方式的不同，将逐线积分方法分为追线积分法和追点积分法。由于逐线积分计算需要耗费大量的计算时间，在大气遥感和大气探测业务中使用时，必须减少计算成本，提高计算速度。本文在追线积分法的基础上，给出了简化的逐线积分的基本方法，在保证同样计算精度的同时，大大提高了计算速度。对在精确的和简化的逐线积分下，不同线翼截断方式(CUTOFF)对吸收系数、大气透过率和冷却率的影响进行了更详细的探讨。通过数值试验发现，对谱线线翼的截断方式是影响辐射计算精度和计算速度的重要因子。在不同压力下，用CUTOFF=2 计算的吸收系数误差最大；对CUTOFF=1，在大多数取样点上误差都小于2%；对CUTOFF=3或4，对绝大多数取样点上计算的吸收系数误差都在5%以内，但所用的计算时间却明显减少。大气低层的透过率对不同的计算方法和不同的线翼截断方式不敏感；对大气高层，无论是对精确的还是简化的逐线积分方法，当CUTOFF=2时的透过率结果与其他线翼截断方式的结果差别较大。通过比较，本文给出线翼截断的优选方案。

英文摘要:

      There are three basic methods in atmospheric radiative transfer calculations, which are line-by-line (LBL) integration, correlated k-distribution method, and band model. The LBL integration is the most accurate one of all. It is divided into two approaches named as integration by lines and by sampling points in the calculated spectral regions according to different integration methods when calculating atmospheric transmissions. Because LBL integration is the most expensive of all, it is necessary and important to deduce calculation time but increase calculation speed when it is put to use in the daily operation in atmospheric remote sensing and atmospheric sounding. A simplified LBL method is given in this paper on the basis of integration by lines, which increases computational speed greatly with keeping the same accuracy. Then, it discusses the effects of different schemes of line-wing cutoff on atmospheric absorption coefficient, transmittance and cooling rate under accurate and simplified LBL methods with a great detail. It is shown by this numerical study that the way to cut spectral line wing has a great effect on the accuracy and speed of radiative calculations. The relative errors of the calculated absorption coefficients are the largest for the method of CUTOFF 2 under different pressures, while for the method of CUTOFF 1, they are less than 2% at most of sampling points and for the method of CUTOFF 3 or 4, they are almost less than 5% in the calculated spectral region, however, the calculation time is deduced apparently. We find in this study that the transmittance at the lower atmosphere is not sensitive to the different approaches for accurate or simplified and the different cutoff methods to the line wing that given in this work. Whereas for the higher atmosphere, the differences of transmittance results between the method OF CUTOFF 2 and other three cutoff methods are the biggest of all no matter for the accurate LBL or for the simplified LBL integrations. By comparison, the best scheme of line-wing cutoff is given in this paper at last.