Incentives for field hydrology and data sharing: collaboration and compensation: reply to “A need for incentivizing field hydrology, especially in an era of open data”*

ABSTRACT

We thank Allen and Berghuijs for continuing the discussion on field hydrology and data sharing and discuss two incentives to promote data collection and sharing in hydrological sciences: a collaborative attitude and additional funding to make data publicly available.

KEYWORDS:

• experimental hydrology
• open-access data
• data sharing
• collaboration
• funding for data publication
• field protocols and standards

Editor A. Castellarin; Associate editor not assigned

We thank Allen and Berghuijs (2018) for contributing to the discussion on field hydrology by commenting on our opinion paper (Blume et al. 2017), raising the important issue of incentives to share data and also for discussing the dilemmas of data sharing.

Data archiving is extremely important because studies have shown that most data are lost within 20 years (Vines et al. 2014). As hydrological systems are changing, it is even more important that the data that are collected now can later be used to assess and quantify these changes (Laudon and Taberman 2016). It is also important that data can almost immediately (i.e. after a reasonable embargo time to allow students to finish their PhD projects) be used by others to increase the data’s potential to advance the hydrological sciences. We think that including data providers in the analyses and as co-authors is the most promising way to improve rapid data sharing because it benefits not only the data users but also the data providers.

While national or regional agencies may, so far, have published most of the publicly available hydrological data (e.g. streamflow and precipitation data from gauging networks), we focus here on data from research projects (including PhD projects). Yet, many of the benefits of collaborating with data collectors and providers are valid for national or regional datasets as well.

The data collectors and providers may benefit from data sharing and collaboration by learning new methods (e.g. models or analysis techniques that they may not be familiar with), new insights, additional publications, expanding their professional network, etc. (see Laudon and Taberman (2016) for a discussion on the value of collaboration). A citation of the data source does not provide the same incentive to the data providers as collaboration because data publications and citations are still not credited similarly or even higher (as they should be) than research articles.

Only by collaborating with the people who know how and where the data were collected and/or generated, know the uncertainties and potentials for errors and how to interpret patterns in the data, can data users be sure that they use the data appropriately and do not misinterpret their results. As experimental work is most often carried out in the field and each field site has its own idiosyncrasies, there are a lot of details and nuances that cannot be put in the meta-data. Single samples or individual measurements are unlikely to be included in the data archive but could, like field observations (or other soft data), be highly relevant for the analyses. Furthermore, the data providers will know much more about potential changes at their research sites (e.g. land use or land management changes that cannot be determined from remote sensing data or land use maps) or measurement set-up and collaborating with them will help to avoid misinterpretation of trends in the data (Whitfield 2012). A collaborative approach will also help to narrow the gap in communication between modellers and experimentalists (Seibert and McDonnell 2002, Fenicia et al. 2008) and thus has the potential to improve both models and experimental design.

For experimental hydrology, there are few established protocols and therefore field equipment may be installed (and used) differently in different studies. As a result the data may not be comparable (e.g. there is no clear agreement on whether or not to purge shallow wells before sampling, or if sapflow sensors need to be de-installed over the winter). However, it is important to note that even laboratory analyses may produce very different results (see the isotope results from Orlowski et al. (2018) for an example). Field or experimental hydrology would benefit from establishing, or being more rigorous in following standard protocols. This would make it possible to assure quality standards with respect to the data and interpretation and would also facilitate site-to-site comparisons. But because experimental work often involves the use of new sensors and equipment, protocols will need to be adapted and published constantly and collaboration with the data providers remains necessary to ensure that the data are used and interpreted correctly.

In the context of data sharing it is also important to acknowledge how much time and effort it takes to write up meta-data and to bring data sets into a shape where others can use them. This is time that is hard to find. Despite time being such a scarce and valuable resource, 80 of the 336 participants of our survey suggested that one should spend however long it takes to make the data available and almost 60 participants suggested that every group that collects field data or runs laboratory experiments should employ someone full time for this task. When we asked how much time they would personally be willing to invest in making their data available, 26% were willing to invest 1–5 work days per year, 40% were willing to invest 6–20 days per year and 23% would be willing to do even more than that (10% of the respondents did not answer this question; Blume et al. 2017). While publishing data “in slices” along with the corresponding research articles is already a major effort, it is even more work to archive entire datasets. Nevertheless, the latter approach is more advantageous for data users because it provides all data from a specific project or field site at a single location and can also include unpublished data that may be useful for further analyses.

PhD students funded on projects are usually under a lot of pressure to finish their thesis before the project money runs out and can rarely allocate the time that is needed to archive and document the data in sufficient detail to make them useful for others. In addition, to be competitive with others, they likely feel pressured to write another research paper instead of spending time on writing the meta-data. The PIs on the other hand are likely already occupied with writing the next proposal or publications for the project, so that data sharing is also less of a priority for them (in addition, they may not know all the ins and outs on how the data were collected). A possible way forward would be to convince funding agencies to provide supplementary funds for projects generating field data to finance the PhD students or technicians for an additional time period, during which the data are processed so that they become available and useful for the scientific community. The Swiss National Science Foundation has started to make these types of funds available and we hope that other funding agencies will follow.

Combining funds for data publication and a general collaborative attitude towards the data collectors could be a beneficial way forward to give field data collection the appreciation it deserves, while at the same time making sure that the data can be used widely and thus contributes more efficiently to the advancement of hydrological science than is currently the case.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

*Allen, S.T. and Berghuijs, W.R., 2018. A need for incentivizing field hydrology, especially in an era of open data: discussion of “The role of experimental work in hydrological sciences – insights from a community survey”, Hydrological Sciences Journal, doi:10.1080/02626667.2018.1495837