Have you ever noticed how companies sometimes attribute step changes in their finances to “accounting procedure adjustments”.

The same may just be true for New Zealand’s glacier volume.

Measuring glacier volume is very difficult, so estimating it using models is a preferred approach. The choice of model affects the result. A new paper by Daniel Farinotti¹ and no fewer than 36 co-authors (including Brian Anderson from Victoria University of Wellington) investigated this model-choice effect. Farintto and friends collected glacier depth observations from 21 glaciers around the world (including New Zealand,s Brewster and Tasman Glaciers). They then sent out a call to glaciologists to model the depths of those glaciers, but he didn’t give them the observed data. After the depths were modeled, they compared the results to the observations. 17 models were applied to 21 glaciers. None of the models proved superior to the others, and their skill varied widely, though combining them led to estimates that were on average within 10 % of the observed values, which you could argue, is pretty good.

In New Zealand, the total glacier volume has previously been estimated² at 53.3 km³, 63 km³ and 67 km³. The largest of those estimates was prepared using a “shear-stress-based approach”, and is one of the model-types tested in Farinotti’s experiment. For the Tasman glacier, this model type, and the average of all the models in the experiment, underestimated the Tasman’s depth. As the Tasman dominates the glacier ice volume of the Southern Alps, the results of Farinotti’s work indicate we have more ice in New Zealand’s glaciers than previously estimated.

Thanks to Farinotti, an accounting procedure may have just given New Zealand a glacial advance!