Development of Spatially Varying Parameters for Arctic Landfast Ice
Abstract
This report describes the development of spatially varying landfast ice parameters and a variational algorithm for the optimization of these coefficients for the Arctic Ocean into the Community Sea Ice Model (CICE6). The CICE6 model code was modified in the generalized form and can be used for incorporation of spatial variability of other CICE6 model parameters. For the variational optimization of the landfast ice parameters we adopted an iterative Conjugate Gradient (CG) algorithm, which requires the knowledge of the cost function and its gradient on each iterative step. In order to increase the computational efficiency we utilized the reduced space approach, where the limited number of the coefficients define the landfast ice in the large area. In addition, the entire Arctic Ocean was subdivided into 9 regions, where landfast ice was observed during the last two decades. The optimization of the landfast ice parameter was conducted separately for each region through the minimization of the difference between observed and CICE6 model generated landfast area. Comparisons are made between modeled landfast ice and observations for 9 Arctic regions.
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 21, 2023
- Accession Number
- AD1211281
Entities
People
- David Hebert
- Elizabeth Douglass
- Gleb Panteleev
- Richard Allard
Organizations
- United States Naval Research Laboratory