What's new in Java OceanAtlas 4.0
This is perhaps the most major upgrade to this program since it
was initially developed. Along with many quality fixes there are many
new features and enhancements to the user interface. The most notable
internal change has been support for the Java 2 Graphics
platform. Java 2 graphics give JOA access to multiple, high quality
fonts, anti-aliased text and graphics, true rotated text, and superior
printing. All of JOA's plots are much closer to "presentation quality"
when printed.
New or Improved Features in JOA 4.0
Spreadsheet File Improvements
- Argo Import: Fixes/changes to spreadsheet file
reader to accommodate Argo drifter data files created by NODC (.jos
files):
- Date and Time can now be in separate columns
- Station observations are automatically sorted by pressure (as
needed) to be in order of increasing pressure.
- Export: When exporting to a spreadsheet
file, one can now select from a variety of built-in and
user-definable missing value indicators and field
delimiters. Missing values can be -99 (JOA default), -999.0 (WOCE
default), 1e35 (EPIC default), space character, or a user-defined
value. Field delimiters can be tab (JOA default), comma, space
character, or user-defined.
- Import Preview: Opening a spreadsheet file (.jos),
presents a new window to preview the spreadsheet file and select
missing values and field delimiters. JOA displays the first 10 lines
of the file in a scrolling text field.
- Import: JOA now accepts -99, -999.0, 1e35, the space
character, and a user-defined string as missing value
indicators. JOA now accepts tab, comma, and user defined (non-space)
field delimiters. Note: space can be used as an output field
delimiter but not as an input delimiter.
Metadata Color bars
Color bars can be created for four common types of station
metadata: date/time, month, latitude, and longitude. The color bar
creation dialog has been modified to accommodate these new types of
color bars. Metadata color bars are dynamic; they do not correspond
to a specific range but are scaled according to the metadata ranges
of the currently open data files. Adding data to an existing data
window will cause all metadata ranges to be recomputed and any plots
colored by metadata to be redrawn. Metadata colorbars can be applied
to property-property plots, profile plots, and map plots.
Interactive Data Enhancement/Filtering
A range of the color parameter can be highlighted on
property-property and profile plots to highlight the points that fall
in that range of the color parameter. We call this feature
ÒenhancementÓ but it is really a form of an observation filter. A
"real" observation filter (i.e., one created through the observation
filter dialog) will take precedence over an enhancement. Enhanced
values are drawn on top of other values for better visibility. There
is a new preference panel for setting the behavior of plot
enhancement.
Map Improvements
JOA maps are greatly enhanced in this version. Improvements
include:
- Presentation quality improvements: Maps have
"cartographic" borders and graticule labels.
- Map Station Symbol Coloring Enhancements: Map plots have major
new options for coloring station symbols including coloring station
locations by station calculations values (e.g., depth of mixed
layer, or value interpolated onto a neutral surface) or by
metadata. Station coloring by isosurface value can now interpolate
missing values of an isosurface.
- Map Improvements for Float Data: Other map
improvements to accommodate drifter data (smarter station connection
line) and bug fixes.
- Station labels: Maps can show the station number or the
value of a station calculation parameter or station metadata next to
a station symbol. User has control of label size, offset, angle, and
display precision.
Color Bar Improvements
Color bar legends now are sized dynamically--they will resize as
one resizes windows with color bar legends. The color bar editor has
been enhanced to allow creation of metadata colorbars and easier
editing of existing colorbars.
Data Window Cross-section Improvements
Cross-sections can now be zoomed by click-dragging in the depth
dimension or setting explicit ranges in its "Properties" dialog. Click
browsing is now more accurate in selecting the closest observation to
the click point. Observation symbols can now be colored by
presence/absence of a specific parameter (in addition to quality
code). User interface has been improved to allow more than 6 station
calculations.
Station Calculations
This area of JOA has changed dramatically. In addition to a new
user interface, there are these new calculations:
- Neutral Surface: Salinity, temperature, and pressure
(and optional error terms) can be interpolated onto user-defined
values of gamma.
- Station Extrema: Calculate the extrema (max, min) of a
parameter over a whole profile or in a user-defined range of another
parameter (e.g., max of O2 between 1000 and 2000 decibars). Can return
the values of other parameters at the extremum.
- Interpolation: Calculate the value of a parameter
interpolated onto the value of a surface (e.g., salinity on a density
surface). If pressure is the surface variable, can optionally
interpolate to the surface or bottom.
Results of all station calculations can be plotted as a station line
plots or used to color map symbols on a map plot. Exporting station
calculation values now allows you to select a delimiter and a missing
value indicator. Exported station calculation files now include fields
for latitude, longitude, and date for better compatibility with
external analysis/plotting tools.
Enhanced Section Manager
- Custom mean Casts: Mean casts
are interpolated onto a user-selectable interpolation surface and
displayed in a new data window. Mean casts can be saved in any of
JOA's output formats. Custom mean casts can be used to create residual
contour plots (described below). Custom mean casts can optionally
include computed statistics, n and variance (in addition to the
mean).
- Add Ocean Bottom: JOA can add the depth of the ocean at
a station's location by interpolation from an etopo bathymetry
file.
- Merge Stations: An option has been added to merge all
stations from all sections into one new section.
Contour Plot Improvements
This is another are area of major improvement in JOA 4.0:
- Residual Contour Plots with Custom Mean Casts: Residual
contour plots can be based upon custom mean casts files created from
other data collections with the Section Manager. When creating a
residual contour plot, JOA allows one to select a mean cast from any
open data window in addition to using a mean cast calculated from the
stations in the current data window. JOA will automatically regrid a
mean cast if it is not on the same grid as the residual contour plot
one has specified. For example, you can compute a mean cast for a
large number of stations in the Pacific Ocean and then plot a contour
plot of residuals for a WOCE section line from this basin-mean
cast.
- Improved Interpolation: Improvements to the built-in
interpolators produce better contour plots. We have renamed and
improved how JOA deals with missing values when interpolating
observations onto standard levels. Note: JOA interpolates to standard
levels one station at a time. JOA now includes three options for
dealing with missing values:
- No interpolation of missing values
- Vertical or "local" interpolation: In this case JOA
attempts to fill missing values by looking a user-selectable maximum
number of observations above or below (depending upon the
interpolation direction) the standard level for a non-missing
value. JOA does not use values from surrounding casts to fill missing
values.
- Horizontal or "far-field" interpolation: Before
interpolating a standard level, JOA will compute for each cast a set
of "virtual" bottles based upon the values of surrounding casts. This
is done by connecting bounding observations by a line, computing the
horizontal gradient, and interpolating where that line intersects the
current cast. The number of surrounding stations to use when computing
virtual bottles is a user-settable option (defaults to +- 2
stations). Virtual bottles are weighted by their distance from the
cast being interpolated. The user also sets how far from a standard
level to look for a real observation. When interpolating a standard
level, if a real observation is not found within n standard levels
(defaults to 2 standard levels) above/below this level, JOA will use
the virtual bottle within this range with the highest weight
value. This technique is particularly valuable when there are stations
with sparse sampling intermingled with more densely sampled
stations
- Edge Processing in Interpolations: The JOA interpolator
includes an option to extend interpolations to the reported ocean
bottom by extrapolating the vertical gradient to the bottom and then
computing horizontal gradients. This produces a much more pleasing
contour plot along the bottom but must be used with caution as false
extrema can be created in certain circumstances.
- Plot symbols for missing values: JOA no longer plots an
observation symbol if the contoured parameter is missing at that
observation (bottle).
- Plot Symbol Color: User can now select the color of
observation/level symbols.
Other Plot Improvements
- Plot toolbars: All JOA plots now have a toolbar at the
top of the plot window. The number of capabilities available in the
toolbar varies by plot type. All plots have a lock button for
"locking" a plot. A locked plot does not allow browsing, resizing,
adding data, or filtering. With the exception of contour plots, all
toolbars include a value browser that shows the value of the plotted
parameters at the cursor location.
- Constrained zooming in property-property plots: Holding
the shift key down when zooming a property plot will only zoom the
plot in the x-axis parameter.
- Multiple browsing cursors in mutiple-axes property-property
plots
- Improved click browsing on property-property plots with
multiple x axes. Click browsing in multiple axes plots now searches for
the closest matching observation for the primary x axis. The primary x
axis can be changed using a plot's settings dialog.
- Isopycnal contours and labels have better presentation
quality appearance Contour lines are continuous and plot labels
are antialiased.
New Preferences
- New Font Preferences: Allows user to select fonts, text
styles, text size, and text color for labels used in JOA
plots. Currently can set styles for: axis values, axis labels,
isopycnal labels, plot titles (printed plots only), colorbar labels,
map graticule labels, and map station labels. Text styles can be saved
in preferences and independent style files.
- Default Autoscale Color Scheme: select which color
scheme to use when creating autoscaled colorbars.
- User-defined browsing cursor: User can select symbol
style (cross, circle, box, etc) and size of linked browsing cursor in
JOA Preferences.
- Better Text Metadata Display: User-settable date and
location formats in Preferences dialog. User can select between a
variety of different date formats and between decimal or
degrees-minutes-seconds format for locations. Formats are used
wherever dates and/or locations are presented (data window, maps,
latitude axes, metadata colorbars, etc...).
Other Improvements
- Improved Page Setup: New Page Setup command for
selecting printer settings.
- Edit Plot Settings Command: Because of a shortcoming in
Java, right clicking (ctrl-clicking on the Mac) to display a plots
contextual menu will often move the browsing cursor. There is now a
Edit Plot Settings command in the Edit menu for displaying a plot's
setup dialog without using the mouse.
- NODC Argo netCDF Data files: JOA can read netCDF files that
adhere to the NODC Argo convention.
- Mac OS X Look and Feel: JOA adheres closely to user
interface conventions for Mac OS X with respect to menu layout and
general look and feel. JOA is integrated with the Finder and supports
custom document types (.joa extension), drag and drop of data files to
JOA's icon open file and launches JOA if necessary, and double
clicking a document launches JOA and opens document.