Minimal glacier models

Research output: Book/ReportBookAcademic

Abstract

During the past few decades the study of glaciers and their response to climate change has shown a strong development. Early theoretical work in the fifties and sixties has been complemented by the construction of numerical models of glaciers and ice sheets with various degrees of complexity. Observations of glaciers from space have given us a much better view on where the glaciers are and what they look like. More recently, advanced techniques like SAR interferometry have revealed an unexpectedly rich dynamic behaviour of many ice bodies. Fast flow (intermittent or continuous) has been observed on a large number of glaciers, and this finding suggests that many glaciers may adjust to climate change more quickly than hitherto assumed. In view of these developments it may seem a bit odd to write a text on “Minimal Glacier Models“. Why play around with simple quasi-analytical models if computer power is virtually unlimited to run codes on very high resolution? Well, digging into to numerical models and codes for more than 20 years has revealed to me that numerical modelling also has negative sides. First of all, documentation of codes (including my own) is normally inadequate, details of how boundary conditions are implemented in a numerical scheme are generally not described, and altogether reproducibility is poor. An outstanding problem is the dependence of numerical solutions on the resolution of a computational mesh or grid, in particular when the balance rate depends on surface elevation, or when the position of the grounding line plays an important role in the evolution of an ice sheet. This criticism is not meant to claim that one should stick with simpler models, but it helps to appreciate that there are more lines along which glaciers can be studied. Simple models are first of all learning tools. There is an intrinsic value in models that can be solved analytically or coded by a student in a couple of days. Simple models have the advantage that the parameter space can be explored in great detail. This means that a better feeling can be developed for what matters and what is fairly unsignificant. Simple models are also very useful to find out how robust the dynamic behaviour of complex numerical models is. Subtle instabilities may be lost when going from many degrees of freedom (in a numerical model with spatial resolution) to a few (in a minimal model). It is useful to have knowledge about the “dimensionality“ of a particular process. In this text the focus is on the interaction between glaciers and climate - not on the mechanics of glacier flow. This interaction involves a number of feedbacks mechanism that can be captured by simple models. The dependence of the balance rate on surface altitude is the most obvious example, and leads to nonlinear behaviour that can be studied well with simple models.
Original languageUndefined/Unknown
Place of PublicationUtrecht
PublisherIgitur, Utrecht Publishing & Archiving Services
Number of pages91
ISBN (Print)978-90-6701-022-1
Publication statusPublished - 2008

Cite this