 02 Jun, 2016 6 commits


Pierre Cazenave authored
Add script to generate a logo based on the MATLAB logo, but with triangulation instead of regularly gridded data.

Pierre Cazenave authored

Pierre Cazenave authored

Pierre Cazenave authored

Pierre Cazenave authored

Pierre Cazenave authored
Add new test for writing out the temperature and salinity at the open boundaries using synthetic data.

 25 May, 2016 9 commits


Pierre Cazenave authored
Remove the reads of the ASCII configuration files (which is very slow) and instead extracted the relevant information from the supplied mesh object. As such, the requirements for the mesh object have changed, so hopefully this won't bite too many people in the behind. Also simplify the allocation of the arrays when uniform values are given (i.e. when in_salt and in_temp are scalars).

Pierre Cazenave authored

Pierre Cazenave authored

Pierre Cazenave authored

Pierre Cazenave authored

Pierre Cazenave authored
Make the sigma distributions be spatially resolved for the UNIFORM, GEOMETRIC and GENERALIZED cases. Also remove the UNIFORM distribution from the checks on the values of parameters which only belong in the GENERALIZED case.

Pierre Cazenave authored

Pierre Cazenave authored

Pierre Cazenave authored
Move the removal of invalid coastline nodes into the function to read coastline nodes rather than having it in the splitting function.

 11 May, 2016 5 commits


Pierre Cazenave authored

Pierre Cazenave authored
Add new function to find the deepest river node within the search radius to improve model stability (by having rivers discharge into as much water as possible. This mostly improves stability for the 'calculated' river inputs).

Pierre Cazenave authored
Make the function have subfunctions for each component of the optimisation. This means if we decide to remove some optimisations, we can more easily see where the logical breaks in the code exist.

Pierre Cazenave authored

Pierre Cazenave authored

 20 Apr, 2016 6 commits


Pierre Cazenave authored
commented out code from the Coriolis function.

Pierre Cazenave authored
Add support for the existing behaviour (where each open boundary could be given a common sponge coefficient for all its nodes) and Rory's addition where every node in a boundary can be given both varying radius and coefficient at each node.

Pierre Cazenave authored

Pierre Cazenave authored

Pierre Cazenave authored

Pierre Cazenave authored

 01 Apr, 2016 5 commits


Pierre Cazenave authored

Pierre Cazenave authored

Pierre Cazenave authored

Pierre Cazenave authored

Pierre Cazenave authored

 29 Mar, 2016 4 commits


Pierre Cazenave authored
Make the returned river names the right shape for later concatenation when filtering out duplicate river nodes.

Pierre Cazenave authored

Pierre Cazenave authored

Pierre Cazenave authored
Extrapolate by removing NaNs from the input data and setting the 'nearest' flag on scatteredInterpolant.

 24 Mar, 2016 2 commits


Pierre Cazenave authored
At the moment those arrays are hard coded, but they could eventually be supplied as an argument to the function. This functionality now means we can scale the physics data (mainl flux, but also temperature and salinity) and the ERSEM variables if we're using FVCOMERSEM.

Pierre Cazenave authored

 23 Mar, 2016 1 commit


Pierre Cazenave authored

 22 Mar, 2016 1 commit


Pierre Cazenave authored
The NEMO ERSEM river data are already in units per metre cubed, so remove that conversion from here.

 18 Mar, 2016 1 commit


Pierre Cazenave authored
