adokter/bioRad: bioRad 0.5.0

bioRad 0.5.0

New CRAN release. All issues included in this release can be found here.

New features

• integrate_to_ppi() is a new function to estimate spatial images of vertically integrated density and reflectivity. This function produces an ppi image showing the density of animals on the earth's surface, corrected for the changing overlap between the radar beams and animal layer with distance from the radar. See Kranstauber et al. (2020, in press) for methodology and this vignette for examples.

• apply_mistnet() is a new function to apply the convolution neural network "MistNet" on pvolfiles to separate biological and meteorological signals (see Lin et al. 2019). Results can be readily visualized with plot.ppi() and plot.scan(). MistNet is now also a segmentation option in calculate_vp() (#262).

• read_pvolfile() and calculate_vp() now read Vaisala IRIS RAW format directly, helpful for countries like 🇨🇦🇫🇮🇨🇴🇵🇹 (#222). bioRad now also reads files containing single elevation scans and calculate_vp() can calculate profiles from multiple files containing single elevation scans (#221).

Newer/faster conversions

• calculate_param() is a new function to calculate parameters from existing parameters, e.g. reflectivity eta from reflectivity factor DBZH (#287).

• scan_to_raster() is a new function to convert a scan into a RasterBrick compatible with package raster (#238).

• scan_to_spatial() is a new function to convert a scan into a SpatialPointsDataFrame compatible with package sp (#238).

• project_as_ppi() is now much faster (e420e5d) and accounts for earth's curvature (820e85f).

New functions describing the radar beam geometry

• beam_distance() and beam_range() are new functions to relate range (i.e. slant range), distance (i.e. down range) and height of the radar beam.

• beam_profile() is a new function to calculate for a set of beam elevations the altitudinal normalized distribution of radiated energy by those beams.

• beam_profile_overlap() is a new function to calculate the distribution overlap (in terms of Bhattacharyya distance) between a vertical profile (vp) and the vertical radiation profile of a set of emitted radar beams (given by beam_profile()).

Additional features

• custom color scales in map() and plot.ppi (#318)

• nyquist_velocity() is a new function to calculate the unambiguous velocity of Doppler radar from its pulse repetition frequency/frequencies (#208).

• filter_vpts() is a new function simplifying the selection of time ranges and instances in vertical profile time series (vpts) (#241).

• The definition of height-integrated velocity is now corrected (#232, #233, 72be6d1).

• Improved documentation of how mtr can be calculated from vid, u, v (6dce625).

• read_pvolfile() now also reads quantities DBZ, TH, T, because these often occur in European data and are relevant for biological analysis (note that DBZ and T are not ODIM-complient names) (5db08bd).

• plot.scan() now has correct ordering of rays (#285).

• plot.vpts() now has correctly positioned and rounded speed barbs (#277, #244).

• integrate_profile()'s default unit of interval argument has been changed to seconds (#234).

• project_as_ppi()'s default projection is now on earth's surface (#280).

• height is now the default quantity denoting height above mean sea level. HGHT is deprecated (#273).

• plot.vpts() and plot.vp() plots are shifted up by half the height interval to reflect that height refers to the bottom of the height interval (#277, #198).

• calculate_vp()'s default sd_vvp_threshold parameter value at S-band is now 1 m/s (#93).

• And many small bug fixes and documentation improvements.