Fix for evaluation errors

This release fixes an error where the Evaluation module would not complete execution, because of the use of numpy.ma.compress use when calculating percentiles for spatial variables like track density and genesis density.

Changelog:

• Add scenario modelling example (for tcevent.py);
• Reduce memory overhead when extracting timeseries data from scenario simulations;
• Fix incorrect wind speed calculations arising from spurious pressure rates of change;
• Add Akima interpolation scheme for interpolating single track positions;
• Fix for numpy v1.10.1 changes;
• Remove masked array compress from percentile function in evaluation routines;
• General clean up of code;
• Licencing and contribution information updated.

TCRM version 1.0

This is the first release of the Tropical Cyclone Risk Model, for estimating the wind hazard from tropical cyclones. The model is under active development by Geoscience Australia.

A benchmark event set generated with this version of the model is available through the National Computational Infrastructure Data Collections, as part of the Australian Natural Hazards Data Archive, where the data can be accessed via HTTP, OPENDaP, WMS, NCML, NetCDFSubset services and others.