Forcing a Numerical Ice Flow Model with Mass Balance Estimates Derived From Tree Rings: Experiments on the Historic Advance of Columbia Glacier, South-Central Alaska
Tidewater glaciers are major contributors to sea level rise, but due to their non-linear relationship with climate variation, their dynamic flux to the world’s oceans is not well-constrained. Prognostic models are therefore crucial for illuminating the controls on their advance and retreat behavior...
| Main Author: | |
|---|---|
| Format: | Text |
| Language: | English |
| Published: |
Open Works
2019
|
| Subjects: | |
| Online Access: | https://openworks.wooster.edu/independentstudy/8476 https://openworks.wooster.edu/cgi/viewcontent.cgi?article=11396&context=independentstudy |
| Summary: | Tidewater glaciers are major contributors to sea level rise, but due to their non-linear relationship with climate variation, their dynamic flux to the world’s oceans is not well-constrained. Prognostic models are therefore crucial for illuminating the controls on their advance and retreat behavior and informing projections of rising seas. Here, a numerical model of ice flow is leveraged to simulate the historic advance of Columbia Glacier, a large tidewater glacier in south-central Alaska. Glacier accumulation begins at high elevations, then the glacier flows downhill into the ocean, where iceberg calving is initiated. A morainal shoal is built and maintained at the terminus, which restricts calving and enables advance into deep water. A new, annually-resolved, multi-century reconstruction of surface mass balance derived from tree rings forces glacial expansion. Exhumed trees overrun by Columbia Glacier constrain the timing of the glacier’s historic advance, offering a series of observed terminus positions with which to validate the modeled terminus positions. In agreement with previous work, predicted terminus positions are relatively insensitive to mass balance variation. A stronger fit between observed and predicted terminus positions is achieved when the shoal volume is decreased as a result of enhanced sediment diffusion along fjord tributaries. However, the model in its present form likely underestimates climate sensitivity to an unknown degree due to artificially low sliding velocities. These results demonstrate the utility of prognostic glaciological models when coupled with extended records of paleoclimate and terminus positions and shed light on the complex dynamics of a consequential glacier. |
|---|