南京师大学报(自然科学版)
南京師大學報(自然科學版)
남경사대학보(자연과학판)
JOURNAL OF NANJING NORMAL UNIVERSITY (NATURAL SCIENCE EDITION)
2010年
1期
112-119
,共8页
黄昌春%李云梅%孙德勇%乐成峰
黃昌春%李雲梅%孫德勇%樂成峰
황창춘%리운매%손덕용%악성봉
漫衰减系数%敏感度%Hydrolight模拟
漫衰減繫數%敏感度%Hydrolight模擬
만쇠감계수%민감도%Hydrolight모의
diffuse attenuation coefficient%sensitivity degree%hydrolight simulation
漫衰减系数K(z)是湖泊光学研究中的重要光学参数,而其中下行漫衰减系数K_d(z)目前的计算方法主要是经验算法,而这些方法对于理解K_d(z)以及影响Kd(z)计算不确定性的分析是远远不够的,本文通过求解辐射传输方程得到K(z)的解析模型,分析固有光学属性(b_b(z))/(a_(z))以及表观光学属性(1)/(u_d(z))和(1)/(u_u(z))对计算K(z)时误差引入的敏感度,根据其敏感度为0的分界点,推演出K_d(z)在(b_b(z))/(a(z))<1和(b_b(z))/(a(z))1条件下的综合分段模型,预测值与Hydrolight模拟值方差为1.0023,标准差为1.0011.
漫衰減繫數K(z)是湖泊光學研究中的重要光學參數,而其中下行漫衰減繫數K_d(z)目前的計算方法主要是經驗算法,而這些方法對于理解K_d(z)以及影響Kd(z)計算不確定性的分析是遠遠不夠的,本文通過求解輻射傳輸方程得到K(z)的解析模型,分析固有光學屬性(b_b(z))/(a_(z))以及錶觀光學屬性(1)/(u_d(z))和(1)/(u_u(z))對計算K(z)時誤差引入的敏感度,根據其敏感度為0的分界點,推縯齣K_d(z)在(b_b(z))/(a(z))<1和(b_b(z))/(a(z))1條件下的綜閤分段模型,預測值與Hydrolight模擬值方差為1.0023,標準差為1.0011.
만쇠감계수K(z)시호박광학연구중적중요광학삼수,이기중하행만쇠감계수K_d(z)목전적계산방법주요시경험산법,이저사방법대우리해K_d(z)이급영향Kd(z)계산불학정성적분석시원원불구적,본문통과구해복사전수방정득도K(z)적해석모형,분석고유광학속성(b_b(z))/(a_(z))이급표관광학속성(1)/(u_d(z))화(1)/(u_u(z))대계산K(z)시오차인입적민감도,근거기민감도위0적분계점,추연출K_d(z)재(b_b(z))/(a(z))<1화(b_b(z))/(a(z))1조건하적종합분단모형,예측치여Hydrolight모의치방차위1.0023,표준차위1.0011.
Diffuse attenuation coefficient K(z) is an important optical parameter in the research of lake optical property, the current calculation methods of downwelling diffuse attenuation coefficient K_d(z) is mainly experience algorithm, and all of these methods for understanding K_d(z) and the uncertainty effects of the K_d(z) analysis are far from enough.By resolving the radiation transfer equation, we got the analytical model of K_d(z), by analysing of the sensitivity degree of the inherent optical properties (b_b(z))/(a(z)) and the apparent optical properties (1)/(u_d(z)),(1)/(u_u(z)) to the introducing error when calculate Kd(z), we deduced the Kd(z)'s Sub-integrated model under the condition of (b_b(z))/(a(z))<1 and (b_b(z))/(a(z))1 according to the zero sensitivity boundaries points, the variance and standard deviation between the Hydrolight simulation and predictive value of model is 1.0023 and 1.0011 respectively.