2023 is the year of open science! Many ongoing programs are taking this year to highlight and help improve the state of open science within different communities. Additionally, NASA has stated that they are committing to this cause long-term. The massive push to have open science become the norm starts this year.
What is "Open Science," you ask? Open Science is the philosophy that through openness and free access to literature, software, data, and other research resources, science becomes more innovative as access to science eases, and the bar of entry for completing science is lowered. Open science practices address many issues, such as needing to be at a wealthy institution that pays for access to top-tiered research journals, having a grant with enough funds to pay to have your research published, travel to give talks, being able to afford or have access to computers and coding software, and many other examples. However, open access is also beneficial to those who practice it. For example, papers published in an open-access form have a measurable increase in citation rates. However, publishing an open-access journal is often very expensive. For example, Nature raised their fees so that the authors now pay £9500.00 to have their journal articles free for people to read. This price is equivalent to what many researchers make in a year in some countries.
Open science, or rather the lack of open science, has a small but personal connection for me. When I was completing my Ph.D. in Australia, we lost access to many of our field's journals for a few months due to the changes in currency exchange rates. As the Australian dollar weakened during the 2008 financial crisis, our Library needed more funds to pay for all the US and European-based journal subscriptions than usual. The exchange rate led to a delay with the Library renewing our subscriptions until the Australian dollar strengthened - which impacted our abilities to do literature reviews and keep up with the newly published science, even if only for a few months. We were lucky. Other institutions less wealthy than my alma mater have to make those choices every year, and many go without access. The ability to travel to talk with other scientists was also tricky. Open science will only solve some of these problems but will help mitigate most of them. Having journal articles open without fees to read (or publish) ensures people worldwide, regardless of their institution or finances, can publish and read about the latest research. Having websites that show new work, new code packages, and new datasets is invaluable when you cannot travel and find out who is doing what and, thus, whom to contact to become more involved and engaged within the research community. Your access to the community impacts your ability to get jobs, be included in grants and missions, advertise your work, and take your research seriously. The further development of tools using non-proprietary code or coding languages has helped to enable more people to get involved!
With this year of open science, we hope to see more of these philosophies adopted and applied to institutions and by individuals.
The goals outlined by the US federal government and its partners this year of open science are:
1: Establish strategic approaches for advancing open science
2: Promote equitable participation in open science through transparency, integrity, and equity of reviews.
3: Account for open science activities in evaluations and incentives
4: Engage underrepresented communities in the advancement of open science and research.
Not all of these goals are achievable by a single individual, but there are ways we can all contribute. One such way is by working with and developing open-source code.
Open source code can take many different shapes. Ideally, the code would be open to the public and accessible for others to contribute and use from the very start of the development. Unfortunately, others wait until the code is in a "releasable" or "workable" version to release the code to the public for use. Applying these best practices ensures that if someone else picks up the code, it can work - perhaps with minimal usability, but ideally easily and without too much effort to set it up.
More and more journals are asking for the code used for the analysis within a publication to be made public in addition to the data. By ensuring both the data and analysis code is made public, science will benefit by being able to reproduce and more easily verify the results independently.
Open code is growing in popularity. However, there are still concerns about ensuring that the code's creators can get credit and recognition for their code development work. Huge strides have been taken to address these concerns. For example, Github projects are now able to be citable. Some journals, such as the journal of open source software, send your code through peer review and review your short description of the code. Additionally, some journals have technical papers where new tools and methods can be published and thus become citable.
Many years ago, Angeline Burrell, Russell Stoneback, Jeff Klenzing, and others started a yearly CEDAR session focused on open-source code packages for ionospheric/ magnetospheric/ atmospheric research. They invited others from the magnetospheric community to participate in one of our joint GEM/CEDAR workshops. These sessions led to a few of us putting together a review article on the current list of Heliophysics python code packages available.
It has been about five years since that original article came out, and a lot has happened since. For example, the PyHC community has formed, and GitHub and other online repositories have become more used within our broader research community - including more open GitHub repositories. Moreover, a special issue in Frontiers in Astronomy and Space Sciences has been published, effectively updating our original paper summarizing all the current open source python packages.
Working with one of these heliophysics python collaborations or getting involved with the NASA TOPS program are great ways to enter the open source/science community. Another is to do this on a smaller scale. For example, make your own GitHub site, upload your analysis code and provide a "read the docs" webpage, provide Jupyter notebooks and examples, and acknowledge that the code is not meant to be ready for use on another's computer/system but give it a DOI and encourage others to use it.
And with that, Here's to the year of open science- Here is my GitHub. As you can see, I am still learning too. Nevertheless, by focusing on open science this year, we will all be pros by 2024.