Why isn't Lee_Kd_PAR * Zeu_lee a constant for all pixels?

[blesht]

Hi.  I create L2 files using l2gen often adding variables of interest to me, such as Zeu_lee and Kd_PAR_lee.  I've checked the source codes (get_Kd.c and photic_depth.c) and Kd_PAR_lee is obtained simply by inverting Zeu_lee.  The relevant lines of Kd_PAR_lee are:

Zphotic_lee(l2rec, &p, Zp);
if (Zp[ip] > 0.0)
Kd[ip] = 1.0/Zp[ip];

That being the case, it would seem to me that their product should be a constant everywhere.  However, when I check values extracted from different locations in a number of images, this is not true.  Here is an R summary of a set of 722 points drawn from a number of different images:

> test_product <- test$Zeu_lee * test$Kd_PAR_lee
> summary(test_product)

   Min. 1st Qu.  Median    Mean 3rd Qu.    Max.
  4.609   5.718   6.103   5.942   6.301   6.595

Am I mistaken in my assumption?   Are these differences too small to be of concern?  Is something done to the values of Kd and/or Zp before they are output as Kd_PAR_lee and Zeu_lee? (I haven't found anything that would suggest that).

Thanks, Barry

[OB.DAAC - SeanBailey]

Barry,

Zeu_lee gives the depth of the 1% light level.
Kd_PAR_lee uses the first optical depth

Sean

[blesht]

Ah (or should that be 'Doh'?)    Thanks, Sean.  So, if instead of using p.prod_ix = -3, zphotic_lee had been called with p.prod_ix = -2, the resultant Kd would be the calculated Kd to the 1% level and then product with Zeu_lee would be a constant.  The output value of Zeu_lee must come from another invocation of zphotic_lee.  The key is the Kd (1/e) is not the same as the Kd (1%).  Got it ( think). 

Barry