Changelog1 1 3

This is the changelog for MeteoIO release 1.1.3

  • global infrastructure

    • the simplified java binding using jnative has been improved and brought up to date
    • spatial interpolations' status information (correlation coefficient, number of stations used, algorithm used) can be retrieved in a string. This makes it much easier for GUIs built on top of MeteoIO to display some quality feedback.
    • the resampling of meteorological fields can now be applied to applications' specific fields, not only to the base set.
  • plugins, filters

    • the accumulation still had some possibly bad behavior. It has been fully rewritten and should now work properly in any situation (up- and down- sampling)
    • a weighted moving average smoothing algorithm has been added
    • the SMET plugin now supports the units offset and multipliers fields. The input files are now given using the "STATION# = xxx" syntax in order to be consistent with other plugins
    • major efforts have been put into the IMIS plugin. It can now transparently build a station from several substations, including custom averaging of their parameters to get the virtual parameter.
  • spatial interpolations

    • the precipitation distribution described in (Magnusson, 2010) has been implemented
    • a new version of the IDW_LAPSE algorithm has been implemented: for each pixel, only the n neighboring stations are taken into account
    • a severe bug when reading the interpolation algorithm's arguments has been found and fixed
    • several interpolation algorithms can now take an optional lapse rate argument and also optionally the "soft" keyword. This brings another possibility of keeping control of situations where the number of stations suddenly goes down, by keeping the same algorithm but swapping to a fixed lapse rate
    • a new projection method has been added, "frac". This considers a given lapse rate as a fractional coefficient to apply to a reference measurement at a reference elevation (for example, 0.05% precipitation increase per meter, starting at the stations' average elevation). Every interpolation algorithm that supports lapse rates should be able to use it, if given the "frac" keyword. This should lead to a future redesign of the spatial interpolations user configuration syntax to accommodate any choice of elevation dependent regression model...
  • documentation

    • multiple documentation updates, including updates to the examples
  • build system

    • the simplified java binding using jnative can now be built on Windows and Linux systems. It is still separate from the global build system (simple makefile).
    • the third party library detection has been much improved. It works very smoothly with recent versions of cmake, but could still be improved for older versions of cmake (including the 2.6 version shipped by Debian)
    • a cmake file for discovering MeteoIO is now provided