Skip to content
Digital All Sky Camera utilities, for U. Alaska Geophysical Institute cameras
Python Shell
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
.github
archive
cal
dascutils
reference
tests
.coveragerc
.flake8
.gitattributes
.gitignore
ConvertDASC_FITS_to_HDF5.py
DownloadDASC.py
LICENSE.txt
PlayMovie.py
PlotProjectedImage.py
README.md
contributors.md
dasc_hist.py
mypy.ini
pyproject.toml
pytest.ini
setup.cfg
setup.py

README.md

image

Actions Status

PyPi versions PyPi Download stats

DASC all-sky camera utilities

Utilities for plotting, saving, analyzing the Poker Flat Research Range Digital All Sky Camera. (Other locations, too). This program handles the corrupted FITS files due to the RAID array failure on 2013 data. The raw data FITS contain one image per file.

DASC projected images

(plot above created with PlotProjectedImage.py)

Install

Most people will find it useful to have the example scripts and the tests built into the Git repo.

git clone https://github.com/space-physics/dascutils

pip install -e dascutils

Usage

To use the API directly, like:

import dascutils as du

data = du.load('tests/PKR_DASC_0558_20151007_082351.743.FITS')

This returns a dictionary of xarray.DataArray, which is like a "smart" Numpy array. The images are index by wavelength if it was specified in the data file, or 'unknown' otherwise. The images are in a 3-D stack: (time, x, y). data.time is the time of each image. also several metadata parameters are included like the location of the camera.

Save the data using lossless compression to HDF5 by

du.save_hdf5(data, "foo.h5")

Now we give several examples.

Download raw DASC files by time

Download Poker Flat Research Range "PKR" October 7, 2015 from 8:23 to 8:54 UTC to ~/data/:

python dascutils/DownloadDASC.py PKR 2015-10-07T08:23 2015-10-07T08:54 ~/data
  • -w four-letter wavelength in nanometers e.g. 0630

As usual, we assume UTC and do NOT specify the timezone.

convert FITS stack to HDF5

It is very tedious to download large amounts of DASC data in single FITS files. We have tried to make this faster by multi-threading the download, but then the FTP server anti-leeching leaves us with broken downloads. As an alternative in general, it's more convenient to have a single HDF5 file for a day rather than 10,000 FITS files. Convert a bunch of FITS files to HDF5 like:

python dascutils/ConvertDASC_FITS_to_HDF5.py ~/data/2015-10-07 ~/data/2015-10-07.h5
  • -t start stop times to convert

Make movies from DASC raw data files

Play movie of all wavelengths in subplots for files in a directory, for example:

python dascutils/PlayMovie.py dascutils/tests/

additional options include:

  • -t specify time limits e.g. -t 2014-01-02T02:30 2014-01-02T02:35
  • -w choose only certain wavelength(s)

Spatial registration (plate scale)

The cal/ directory contains AZ and EL files corresponding to each pixel.

import dascutils as du

data = du.load('tests/', azelfn='cal/PKR_DASC_20110112')

now data includes data variables az and el, same shape as the image(s), along with camera position in lat lon alt_m.

  • Be sure you know if you're using magnetic north or geographic north, or you'll see a rotation by the declination.
  • Note the date in the filename--perhaps the camera was moved since before or long after that date?

Map Projection

A common task in auroral and airglow analysis is to project the image to an imaginary alttiude, that is, as if all the brightness were coming from that altitude. Each wavelength has a distinctive peak emission altitude due to the energies, kinetic reactions and vertical density profiles involved.

It takes considerable time to stretch the image stack pixels to fit a project geographic grid, so this is not enabled by default. It will be enabled if the du.load(..., wavelength_altitude_km=) parameter is used. See PlotProjectedImage.py for an example.

Some analyses can be done far more rapidly by just projecting pixel(s) of interest rather than an entire 3-D image stack.

You can’t perform that action at this time.