WikiWFS

...and the Rhine-Meuse Delta Datasets

Utrecht University holds a long tradition of studying the landscape, geomorphology and subsurface architecture of the Rhine-Meuse Delta (Dutch: Rijn-Maas Delta, RMD).
These days we are moving these datasets into Open Science domains. A challenge to do so is to find a way where we can release and serve datasets as Open Data,
but at the same time can harvest feedback, correct errors, update insights, append new data and push updates.

This is the reason that we are setting up what we call a 'wikiwfs' system. Herein, WIKI stands for 'community-editable' (as in wikipedia) and WFS stands for 'web feature service' (streaming of vector map data with attributes). We began setting it up in 2019, and the web frontend is moving from alpha to beta deployment in May 2021. The idea and intention is that the wikiwfs system is the main place where the faculty shares medium-sized Open Datasets in their most actual form. At the same time, it is to be a place where editors of the datasets take in feedback from users and push through suggested corrections. These users include staff and students, external collaborators, professionals making regular applied use of the data, or enthusiastic citizen scientists.

We will also continue to deposit our datasets in more traditional ways, i.e. as date-stamped static repositories (e.g. RMD basemap 2012, LLG boreholes 2017). We use the RMD datasets to develop the wikiwfs as these are a data lake of diverse contents in terms of formats (Relational Tables, Point, Line, Polygon, Raster and Voxel gridded data) each of them medium-sized (not millions, but many thousands of features). Distributing and mining and validating and improving such data requires multiple ways of interlinkage (on IDs, on coordinates, via URL, API, GIS service), which the wikiwfs system is intended to provide. Our servers could host further datasets of other groups within the Faculty of Geosciences.

LIMITED/DISABLED FUNCTIONALITY

Sorry for the inconvenience. It will be a major functionality update in the end. Finally ready for public beta.

Welcome students!

We welcome participants to the GEO3-4207 course Fieldwork Low Land Genesis of the academic year 2022-2023.
You will make use of our new developed wikiwfs website, notably the LLG/IT borehole data sections. As such you will be our beta testers!

NOTE: Please use the Edge or Chrome web browser when you download data in XML-format for offline use in our LLG2012 software.
Other web browsers Firefox resp. Safari happen to subtly alter to the XML once generated, which makes LLG2012 show zero boreholes resp. crash.

Contents

From the larger suite of Rhine-Meuse Delta datasets, we currently keep live versions of three main data sets in this wikiwfs system:


Relation with earlier RMD dataset versions


Since 1998, Utrecht University has maintained a GIS system storing the Rhine-Meuse delta channel belt geological mapping and dating, to benefit palaeogeographical reconstructions and analysis of landscape evolution of the delta. The channel belt mapping is based on borehole data (Utrecht University; Geological Survey of the Netherlands), Lidar imagery (www.ahn.nl) and sedimentological and geomorphological principles. The age attribution is based on radiocarbon dating, confrontation with archaeological finds, geological cross-cutting principles - supplemented with knowledge of geomorphological processes (time lags) and findings from the network analysis that the piece-wise mapping and cataloging method made possible. For the youngest generations it is also based on historical sources and maps.

The first version of the maps and catalogs was created in 1998-2000 and published in analogue form of a maps + table-book in 2001 (Berendsen & Stouthamer 2001: Palaeogeographical Development of the Rhine-Meuse delta). The digital versions of table catalogues and GIS map data at the time were maintained by individual researchers at Utrecht University (Berendsen resp. Cohen) and derived copies were distributed on CD-ROM. The product found wide usage in various applied research sectors (hydrology, archaeology, construction) in the densely populated delta plain, besides academic use in a series of PhD-research projects (Stouthamer 2001, Cohen 2003, Gouw 2007, Erkens 2009, Hijma 2009, Bos 2010, Van Asselen 2010, Toonen 2013). With that usage also came feedback, and needs to update insights and locally correct mapping and age assignments and wishes to expand the mapping downstream and upstream.

A second version of the maps and catalogs was released in 2012 and published as a digital dataset repository using the portal https://easy.DANS.knaw.nl. It contained many updates to the mapping harvested from the user feedback (NWO grant to Stouthamer & Cohen), and expansion of the mapped area and time depth considered (not only the Holocene delta evolution, but also the Late Pleistocene valley evolution, in cooperation with W.Z. Hoek). The table of 14C dates used grew, as did the catalog of channel belt descriptions. The 2.0 GIS files and catalog tables were distributed through the download facilities provided by the Easy portal ( https://easy.dans.knaw.nl/ui/datasets/id/easy-dataset:52125 ).

Since then, further interim and regionally expanded versions have been created and reposited as datasets (e.g. interim version of RMD mapping as one of many input layers in national archaeological-geological buried-landscape mapping: Cohen et al. 2017; e.g. regional remapping of the Utrecht-sector of the RMD: Van Dinter et al. 2017; the channel belt age in the Meuse valley in Limburg: Woolderink et al. 2018; dating of the beginning of activity of the rivers Lek and Hollandse IJssel: Pierik et al. 2018.). These communications are to be incorporated as part of the 2019-2020 filling of the wikiwfs version of the RMD dataset. Potentially, the WikiWFS system can also host and serve their further digital-map RMD data products of the Van Dinter and Pierik (materials archived on https://easy.DANS.knaw.nl).

The WikiWFS server idea


The present way in which academic research data is pushed to the Open Data domain, is by simply offering a download opportunity of a static archived copy of a dataset. Such a copy is date-stamped at the day it was deposited in the repository and researchers prefer to archive it at a repository that will care for long-term availability of the data, offers a DOI and/or other persistent identifier, is good in also communicating metadata and relevance tags, so that the dataset is findable for new potential users and so on. This works fine for a first release of any dataset, whether small or medium large, and whether of topical or generic nature.

However, soon as a dataset becomes a success and a user community gets used to it and wants parts of it updated... one runs into version management problems, risks diverging versions, creates a 'release the update' bottleneck for the dataset-owner and risks maintenance of the dataset to stop because the people involved change job tasks or employer and so on. At least, this was our experience with Rhine-Meuse Delta databases established in the 1990-2010s, that became a national professional usage success as subsurface data is in growing demand by a diverse usership in the urbanized delta and coastal plain of the Netherlands. Two wishes emerged from this: (1) we wanted to be able to server our datasets as WFS services in addition to Open Data downloads, and (2) we wanted our users to be able to edit (correct, update, append) entries in our databases in a wiki-style. These wishes coalesced with the Open Science / Open Data developments ongoing in academia anyways, and the WikiWFS server idea is the cumulation of this all.

The WikiWFS system has as goal to enable hybrid forms of Live Open Data. With this we mean that besides offering data-stamped copies, the system offers opportunity for Open Collaboration (public co-editing), Restricted Collaboration (members only co-editing) and Pre-release Editing (new dataset prototypes). Such should streamline continued maintenance on datasets after their first public release. It is believed that releasing datasets in community-editable form will be a positive thing for individual researchers who launch a new dataset, as well as for communities that are heavy users of datasets and stakeholders in having them generally available and updated where appropriate. It is easier to make it explicit 'who created the dataset originally', 'at what institute', 'where the most actual version of the dataset is hosted' and 'who are currently caring for the dataset' with a WikiWFS system hosted at the dataset's parent institute up and running - than having to fall back on static copies only. It is also believed that this will reduce often hidden labor-bottlenecks regarding dataset maintenance tasks after release. The WikiWFS system is an infrastructure answer to time-management conflicts between maintaining legacy output (keeping to serve user communities at scientific quality standards) and generating new output (moving the science forward).

A first WikiWFS-prototype was developed in 2019-2020 (UU-ICT research-innovation grant to Cohen et al.) and made use of a WikiMedia installation with plug ins allowing to connect to database server (H.N.C. Cox, K.M. Cohen). The reason to upgrade the 2.x datasets and their desktop-management to this server-based 'Web-3.0' version, was that single-person management of catalogs and GIS files became a bottleneck to keeping the databases actual. The wish was to make it possible to keep a 'live' dataset and allow multiple users ('the user community of the product'). Such a 'live' version would then allow editing, correcting, enriching and expanding the factual data and palaeogeographical reconstructions stored in the RMD datasets (K.M. Cohen, H.A.G. Woolderink, H.J. Pierik). The RMD dataset thus was used to prototype 'WikiWFS server': a web address hosting both the descriptive catalogs and the digital map data in interlinked way. In 2020 and 2021 we continued our activities with internal funding (UU-GEO Dean's policy means grant 2020). The systems runs on a hardware server funded internally (Delta Evolution / Low Land Genesis group investment).

For its integration task, the WikiWFS system requires a software and data architecture on a constellation of virtual servers (web server, database server, file server, GIS servers). The WikiWFS system comprises frontend and backend code that make 'the data lake accessible through web interfaces and over intranet. The system is of generic setup. The data server backend is a fairly standard installation of established software (e.g. PostgreSQL). The website frontend contains customized solutions (client-side and serverside javascript, ajax). We are still working on the WFS/GIS service part of the setup: we aim to integrate it with a scalable ArcGIS Enterprise deployment (medio 2021).



The URL wikiwfs.org is reserved to describe the generic idea and provide information to ICT developers, administrators and system architects.
This wikiwfs.geo.uu.nl server is the original prototype server, currently hosting 'live' RMD datasets and ready to host other Utrecht University Faculty of Geosciences datasets.

What are channel belts?


Channel belts (Dutch: 'beddinggordels') are sedimentary geological features encountered in the subsurface of low lands created by rivers also known as ribbon sands ('zandbanen') and where they occur shallow enough to have surface expression as 'alluvial ridges' ('stroomruggen'). The RMD contains many generations of channel belts, that were formed and abandoned by natural processes and affected by humans in the last two millennia. The channel belts are shallow-subsurface sand bodies which are of groundwater-hydrological importance. The sand bodies tend to be topped by natural levees that are modestly elevated above the wider delta plain and the water in the river, which is of archaeological importance (habitable, navigable).

Selected references

v1.0 / research history

- Berendsen, H.J.A. & Stouthamer, E. (2000: Palaeogeography, Palaeoclimatology, Palaeoecology 161: 311-335) -Berendsen, H.J.A., Cohen, K.M. & Stouthamer, E. (2001: Maps and Cross-sections, Chapter and addendums in Berendsen & Stouthamer, 2001: Palaeogeographic development of the Rhine-Meuse delta, The Netherlands, Assen: Koninklijke Van Gorcum, 268 p.)
- Berendsen, H.J.A., Cohen, K.M. & Stouthamer, E. (2007: International Journal of GIS, 21:589-602) https://doi.org/10.1080/13658810601064918

v1.x interim

- Cohen 2003
- Erkens 2007
- Hijma 2009

v2.0

- K.M. Cohen, E. Stouthamer (2012) VERNIEUWD DIGITAAL BASISBESTAND PALEOGEOGRAFIE VAN DE RIJN-MAAS DELTA. Beknopte toelichting bij het Digitaal Basisbestand Paleogeografie van de Rijn-Maas Delta. Dept. Fysische Geografie. Univ. Utrecht. https://dspace.library.uu.nl/bitstream/handle/1874/319441/2012.pdf?sequence=1&isAllowed=y
- K.M. Cohen, E. Stouthamer, H.J. Pierik, A.H. Geurts (2012) Digitaal Basisbestand Paleogeografie van de Rijn-Maas Delta / Rhine-Meuse Delta Studies’ Digital Basemap for Delta Evolution and Palaeogeography. Dept. Physical Geography. Utrecht University. Digital Dataset. https://doi.org/10.17026/dans-xkk-f29b
- LLG2012 Software - K.P. Volleberg (2008; 2012)

v2.x interim

- H.A.G. Woolderink, K.M. Cohen (2018): Digital Basemap for the Lower Meuse Valley Palaeogeography. DANS. https://doi.org/10.17026/dans-xkk-f29b
- Cohen 2017
- Van Dinter 2017
- H.J. Pierik, K.M. Cohen, E. Stouthamer (2016): A new GIS approach for reconstructing and mapping dynamic late Holocene coastal plain palaeogeography, Geomorphology 270, 55-70, https://doi.org/10.1016/j.geomorph.2016.05.037
- H.J. Pierik, E. Stouthamer, K.M. Cohen (2017): Natural levee evolution in the Rhine-Meuse delta, the Netherlands, during the first millennium CE, Geomorphology 295, 215-234, https://doi.org/10.1016/j.geomorph.2017.07.003
- H.J. Pierik (2017): Geomorphological reconstructions of the natural levee landscape in the first millennium AD of the Rhine-Meuse delta, the Netherlands, https://doi.org/10.17026/dans-zg9-nqfx