Measured experimental data have irreplaceable role hydrology. These data provide the foundation to make mathematical models accurate and useful by providing a possibility to check the model outputs against reality. But the expenses to collect such data are tremendously high … you either need to instrument a natural watershed, keep it in a good condition and wait for natural rainfall events to occur or you construct a rainfall simulator and execute in situ or lab experiments, both of which can be high stuff and time demanding. Collecting a dataset large enough to cover sufficient range of input parameters can be not feasible for a single research team.
The structure of the database can be examined and obtained on it’s github repository: runoffDB @ github
The graphical user interface for comfortable management of the contents can be examined and obtained on it’s github repository: runoffDB user interface @ github
Source of a Python framework for datamining the database content: runoffdb_miner @ github
The author of the whole idea can be contacted via institutional email: jan.devaty(at)fsv.cvut.cz
This database project wants to provide an open platform for collecting and sharing rainfall-runoff related measurements, both natural and artificial in a consistent manner.
Every instrumented watershed and every rainfall simulator is different and so are the methods for data collection, recording, processing and utilization. Mining the data from different sources for comparison or a common purpose can be quite exhausting as all the teams and workers use different software, workflows and structures for storing the data. This database is an attempt to provide a robust structure for storing and sharing experimental data together with its metadata, relationships between data sets and other information about the data collection and preprocessing following the “FAIR Guiding Principles for scientific data management and stewardship”.
Having all the data stored at one place in a consistent manner allows for automated tools to be used to retrieve the data in desired structures needed for a particular utilization e.g. calibration data sheet for a certain model. Once the mining tool is ready the calibration sheet can be automatically regenerated in a snap every time the DB grows.
The database is MySQL based including the metadata related to the measurements themselves, the research teams and rainfall simulators as well as the funding sources. Registred users have the possibility to download XML structured export of the DB contents that can be further processed in ones favorite working environment (R, PHP, Python etc.).
The development has been supported by the research project QK1810341 of Czech National Agricultural Research Agency