Spectral diffuse reflectance Rw(0+,lambda) was calculated using a newly
 rewitten model code (get_Rw0p_lambda_Cnorm_theta.F90) and the plots of 
 Rw(0+,lambda) were generated for near-zero Chloro values and compared with
 Rw(0+,lambda, C=0). Spectrally integrated Aw has not been produced (it's time
 comsuming to plot in log scale).


-Diffuse_Reflect_Rw0+C_0to0.0002.eps
-Diffuse_Reflect_Rw0+C_0.0003to0.02.eps

 show, for zenth angle = 60 deg, the spectral diffuse reflectance 
 Rw(0+,lambda) = Tw*gamma*bb/a in Eq.(5)
 for C = 0,0.00,0.00001,0.00002,0.00003,0.00005,0.0001,0.0002 mg/m^3 amd,
 for C = 0,0.0003,0.0005,0.001,0.002,0.003,0.005,0.01,0.02 mg/m^3.
 The curves with open circles are for C = 0 mg/m^3. The plots are close to
 each other for C < 0.005 in the whole spectral range of 300-750 mm.

-DIFF_Diffuse_Reflect_Rw0+C-Rw0+0_C.0005to.02.eps
-DIFF_Diffuse_Reflect_Rw0+C-Rw0+0_C.00001to.0003.eps

 show, for zenth angle = 60 deg, 
 the difference Rw(0+,lambda,Ci)-Rw(0+,lambda,C=0) 
 for Ci = 0.00001,0.00002,0.00003,0.00005,0.0001,0.0002,0.0003 mg/m^3, and
 for Ci = 0.0005,0.001,0.002,0.003,0.005,0.01,0.02 mg/m^3.
 In the 1st fig., the difference is always positive, 
 Rw(0+,lambda,C=0)< Rw(0+,lambda,C=Ci), and so the diffuse albedo Aw, Eq.(4),
 intergated over lamda = 200-750 nm will be positive regardless of the 
 distribution of the spectral solor irradiance, Fd(0+,lambda) 
 for Ci < ~0.002 mg/m^3.  
  

-sfc_albedo_diff_Aw-Aw0_0-30-60deg_5nm_750mxl_wrong_bbp.eps

 This plot was generated using bbp(lambda) in Eq.(7) that is 
 propotional to lambda, ie, ...log10(C))x(550/lambda)] was 
 written as ...log10(C))/(550/lambda)]. This results in the
 60 deg curve that is more resemble to that in Figure 1(b) when compared to
 the fig sfc_albedo_diff_Aw-Aw0_0-30-60deg_5nm.eps.
 The positive difference for smaller Chloro (< 0.1mg/m^3) disappeared
 for all zenith angles. However, this bbp(lambda) is not correct 
 and this figure does not show the correct surface albedo difference Aw-Aw0.

-get_Rw0p_lambda_Cnorm_theta.tar.gz
 
 includes the fortran code, get_Rw0p_lambda_Cnorm_theta.F90,
 to calculate Rw(0+,lambda) and two parameter tables used in it,
 Table_Smith_5nm_300-750.data and Table_Morel_5nm_300-750.data,
 for C = 0.00,0.02,0.03,0.05,0.10,0.20,0.30,0.50,1.0,2.0,3.0,5.0,10.0,20.0,30.0,
 and for lambda = 300(5)750 nm. The code shows how the case of C = 0. was treated.
 The value n defined in Morel was used [a(lambda)*u_morel = K(lambda)]
 and iteratibly solved. Then u_frouin = 1/u_morel was assumed, and u_frouin was
 used in a(lambda) = K(lambda)/u_frouin [Eq.(8)] to find  a(lambda) in Eq.(5).
 I believe there is no coding error in it, but please let me know 
 if you found any.
 The above files are in http://frouingroup/~kyozo/index.php?dir=Chl_Aw/20jan10/