Color Enhancement Images

The simplest type of image to create is a color enhancement.  From the terminal, type:

cwrender --enhance RhoC_620 example_goci2_reflect_jan_2023.hdf rhoc_620.png

You should have a new PNG file in your directory that looks like this:

The legends on the right indicate that this is a 620 nm reflectance image from the GOCI-II sensor and the context image at the bottom right shows that it covers the Korean Peninsula.  By default cwrender computes statistics from the variable data and linearly scales the data values to colors using a grayscale palette and a minimum/maximum range based on the mean and standard deviation of the data values.  This is a reasonable heuristic that gives us an initial browse image with some detail.   

To revise the image, let's add the following function and range options:

--function gamma --range 0/1

This properly scales the reflectance in the range [0..1] measured by the sensor for the non-linear way that humans perceive light intensity.  Some additional options add coastlines and grid lines:

--coast white --grid white

To combine all these options, enter this on the command line:

cwrender --enhance RhoC_620 --function gamma --range 0/1 --coast white --grid white example_goci2_reflect_jan_2023.hdf rhoc_620_new.png

The resulting image should look like this:

Suppose that we want to show the data at its full resolution, ie: 1:1 — we can add a magnification option that specifies the latitude, longitude, and magnification factor:

--magnify 33.4/126.5/1

The magnification factor specifies how large each data pixel should by rendered in the resulting image, ie: the ratio of image to data pixels.  For example, a factor of 1 shows the data at full resolution, 2 shows the data zoomed in by 2x, 0.5 shows the data zoomed out by 2x, etc.  Without any magnification option at all, cwrender fits all of the data into the output image.  Adding the option above will create this image:

We'll give several examples now of cwrender command lines for creating color enhancement images from other data files, and general option recommendations in Table 3-2.

Example 1 — Sea surface temperature in the Red Sea

cwrender --units celsius --enhance sea_surface_temperature --palette HSL256 --range 25/31 --magnify 28/34/1 --coast black/gray40 --grid white example_avhrr_swath_sst_sep_2022.nc sst_red_sea.png

Example 2 — Global chlorophyll-a data

cwrender --enhance chl_oci --palette NCCOS-chla --function log --range 0.01/64 --coast black/gray40 --grid white example_viirs_chlor_sep_2022.nc chlora_global.png

Example 3 — Sea level anomaly in the North Pacific

cwrender --magnify 33/146/2 --enhance sla --palette Ocean-balance --range -1.5/1.5 --coast black/gray40 --grid white example_altim_surface_curr_feb_2023.nc north_pacific_sla.png

Bonus exercises:

• Look at the manual page for cwrender and discover what other overlay options are available, for example for political lines, bathymetric contours, and expression masks.  For example, try using --political red .
• Try using the --size option to alter the output image size — the default is 512 pixels.
• Change the output file name to end in .pdf, .jpg, or .tif and inspect the output.  Notice the difference in output file sizes.
• Try running with --nolegends --size full to render just the data with no legends at full resolution. 
• Run cwrender with just --palettelist and no other options or parameters to see what it prints.
• To find out what options were used with cwrender to create an output image, try this at the command line:
  strings image.png | grep cwrender (Unix)
  format-hex image.png | more (Windows PowerShell)