Zooming out from small patches and watersheds to regions: what are missing out in the modeling world?

Large-scale ecosystem and land-surface modelers often get stuck in a hard choice: bottom-up approach, e.g. from the single leaf to single tree, from single tree to forests; or top-down approach, e.g. see forests as forests that are composed of many trees. Whatever the choice is, one more major decision can’t be voided: how big the indivisible element or generalization is enough. This presentation shows a case study over a mountainous region in the Pacific Northwest to see how information could be lost with a bottom-up approach when zooming out from patches to a region scale by using different scales.