v0.100.1

Oceananigans v0.100.1

Diff since v0.100.0

Merged pull requests:

• Eliminate GridMetricOperation and create node metrics (#4805) (@glwagner)
• (0.100.1) Set CUDA.jl compatibility to v5.9+ (#4824) (@ali-ramadhan)
• Include AbstractOperations in distributed reductions (#4829) (@simone-silvestri)
• Minor change in regridder interface (#4830) (@simone-silvestri)
• Support time stepping with DateTime clock (#4831) (@glwagner)
• Simplify README header (#4832) (@glwagner)

Closed issues:

• "Number of threads per block exceeds kernel limit" when computing a reduction on a KernelFunctionOperation on some GPUs (#4047)
• Calculating cell time scale in triply bounded ImmersedBoundaryGrid errors on GPU (#4663)
• ERROR: Number of threads per block exceeds kernel limit (640 > 512). for small immersed-boundary problems with TimeStepWizard and PartialCellBottom (#4694)