Tomoe Note 003: Checking the "weather"

Michael Richmond
Apr 17, 2017

When one analyzes a set of data from Tomoe, one occasionally finds strange or unexpected events. Are they real? Well, sometimes, strange results can be caused by what I'll call bad weather -- observing when

• the sky is very bright (due to moonlight or clouds)
• stars appear fainter than usual (due to clouds)
• the FWHM is large
• the PSF changes rapidly in shape or size with time (high wind)

During periods of "bad weather", we might expect some real objects to disappear, even if their luminosity is not really changing.

In order to help us understand the data from Tomoe, I have created a tool which uses information output by the Tomoe pipeline to create a single large graphic covering some space of time, up to a single night; the graphic shows several indicators of the data quality, acting like a "weather report."

Below is an example, showing properties of some data taken on Apr 11, 2016. The dataset is the composite FITS file TMPM0109330.fits, which contains 360 Tomoe images taken during a period centered on JD 2457490.1051285 = 2016 April 11 14:31:23. The field location is

  (J2000)   RA = 196.5758   Dec= -16.2223

The conclusion one can easily draw from this particular set of data is that the conditions improved from the start of the run to the end, as the following explanations should make clear.

The panels show

top panel: number of images

Each "chunk" of Tomoe data contains a large number of individual FITS images, bundled into a giant FITS file with 3 dimensions. The number of images is currently 360 per chunk. This panel shows the number of individual images within each chunk which pass certain tests during the pipeline processing:

• raw is the number of raw images
• ens is the number of images with enough stars to become a member of the photometric ensemble for this chunk
• ast is the number of images in which astrometric calibration has succeeded
• phot is the number of images in which photometric calibration has succeeded

In this panel, a large value is GOOD.

second panel: zeropoint

The "zero-point" is the difference between the instrumental aperture magnitudes produced by the pipeline and the calibrated V-band magnitudes based on a comparison with UCAC4 catalog. The sense is

                zero_point = (intrumental_aper_mag) - (UCAC4_V_mag)

           

So, larger numbers (more positive) mean "lots of clouds", and smaller numbers (more negative) mean "more clear."

In this panel, a large number is BAD.

third panel: sky value

The background sky value (in counts) determined around the stars in each chunk are averaged to yield a single "sky value" for the chunk.

In this panel, a large number is BAD.

fourth panel: FWHM value

The Full-Width at Half-Maximum for each star is averaged for the entire chunk, and converted to units of arcseconds.

In this panel, a large number is BAD.

fifth panel: total number of stars

The number of stars detected in each image in a chunk is added together to compute the total number of stars found in the chunk.

In this panel, a large number is GOOD.