Dear Iolite team, (sorry this is long)
I recently U-Pb dated some 3 Ma zircons (very low 206 and 207Pb) and made Tera-Wasserburg plots in IsoplotR. I noticed a huge improvement if I make the TW plot using the 207/235 and 206/238 ratios, versus if I plot using the 238/206 and 207/206 ratios (see figures). Note that these two plots show data from Iolite for the same 44 analyses.
I exported my data as time series and recreated all of Iolite's calculations. I realized the data plotting in negative TW space result from my counts on 206 dropping below baseline at some points during the 30 second ablation. This causes the 206/238 ratio for those points to approach zero, so that when Iolite calculates the mean of the ratios it gives a reasonable result. But when 206 drops slightly below zero the calculated 238/206 ratio becomes a huge negative number that severely skews Iolite's mean of ratios calculation, causing the Final238/206 value to be negative. The same is true for 207/206.
For my example, each analysis includes 150 points for a 30 sec ablation. I have the Iolite normal stats set to mean with 3sd outlier rejection and I also tried 2sd outlier rejection, but none of my 238/206 points were rejected to my surprise. In one analysis there were 16 points that were very negative, with 238/206 values around -25,000 to -40,000. The mean value of these 150 points, as given as the Final238/206 by Iolite, is -550.
However, if I copy the Final207/235 and Final206/238 ratios from Iolite into IsoplotR in order to create a TW plot, IsoplotR clearly just calculates 238/206 as 1/(206/238), which gives a nice positive value. IsoplotR likewise apparently calculates 207/206 as ((207/235)/(206/238))/137.818, again giving a nice positive value. But is this a valid approach? Surely 238/206 and 207/206 ratios should be used for plotting in TW space? Ideally it seems like Iolite could reject ratios that are large negative numbers. Please make suggestions if I should make some other change in my data workflow.
thank you,
Lisa