-
Notifications
You must be signed in to change notification settings - Fork 1
Experiments
Ice calving can take many forms, from the calving of large, floating ice shelves in Antarctica to marine terminating glaciers in Greenland. When calving occurs in nature it tends to happen at discrete intervals, however when considering the long-term evolution of ice sheets we are often times more interested in the long term trends of these individual calving events averaged out in time. For the sake of clarity, we hereby define the terms discontinuous calving to refer to an individual calving event and continuous calving to refer to the long term average of discontinuous calving events, usually expressed as a calving rate in horizontal distance of ice calved over time. The relative difference between this calving rate and the ice velocity at the calving front provides the change in calving front position, with an ice front experiencing no calving advancing in line with ice front velocity.
We also make a distinction between a calving law and a calving algorithm. A calving law is a representation of the physical processes behind ice shelf calving used to determine when and how much ice should calve. A calving algorithm, on the other hand, is the implementation of calving law within a numerical model. The implementation of a calving algorithm within a numerical model should be independent of a particular calving law.
Current large-scale, long-term simulations of future ice sheet predictions have (with a few exceptions) tended to ignore the effect of ice shelf calving, either maintaining a fixed, unmoving calving front position or allowing the ice to thin to a minimum thickness due to increased oceanic melting. There is the potential for these simulations to be underestimating future mass loss from ice sheets due to the inability to represent the physical process of calving.
CalvingMIP seeks to address this knowledge gap through three distinct phases of experiments. Phase one experiments are intended to act as a capability check on numerical models calving algorithms, phase two experiments are intended to a comparison between various calving laws and numerical models using an idealised domain whilst phase three experiments will use real world cases to examine to perform future simulations of calving.
Through out all CalvingMIP experiments we require participants to use the following assumptions and parameter values unless otherwise stated in individual experiment discriptions.
These three phases of experiments currently have a focus on continuous calving, with relatively large domains (approximately 2000 square kilometres over a period of hundreds of years) and as such may not be well suited for models that are simulating discontinuous calving at high temporal and spatial resolutions. If there are a particular set of experiments that would better suit this style of calving we would be agreeable to adding in another phase of CalvingMIP specifically looking at this type of calving.