Institutional research project · 2025–2029
GEOSKLAD
A concept for a geospatial data warehouse to support environmental decision-making
Earth observation produces far more environmental data than anyone currently uses. GEOSKLAD works out how to store and structure it so that decisions about the environment can actually be made from it.
EN | HR
48 months
November 2025 – September 2029
4
environmental themes
5
study areas across Croatia
7
collaborating institutions
Host
University of Zagreb, Faculty of Geodesy
About the project
Satellites, ground sensors, weather stations and public registers generate enormous volumes of data about the environment. Most of it is never reused, because it sits in incompatible formats across sources that were never designed to talk to each other. GEOSKLAD asks what it would take to bring that data together in one place.
Over four years the project studies four environmental phenomena in Croatia and treats them as a working test set: urban heat islands and urban vegetation in Zagreb and Split, water quality in Lake Vrana near Biograd na Moru, and landslides at Šandrovec and Kutina. The question is not only what each dataset shows on its own, but how the different data types relate to one another, and what structure would let a decision-support system draw on all of them at once.
The main result is a blueprint — a documented concept for a geospatial data warehouse — together with the analyses, atlases and visualisation methods produced along the way, which remain available to other researchers after the project ends.
Objectives
01
A concept for the data warehouse
Define what a geospatial data warehouse for environmental decision support has to hold, and how it should be structured. Everything else in the project feeds into this.
02
Urban heat and urban vegetation
Map heat islands and vegetation cover in Zagreb and Split from satellite imagery, from May to September, across the life of the project.
03
Lake water quality
Estimate optical water-quality parameters for Lake Vrana near Biograd na Moru from satellite imagery, year-round rather than seasonally.
04
Landslides and erosion
Survey landslides at Šandrovec and Kutina with photogrammetry and remote sensing, supported by field campaigns twice a year.
05
Visualisation and result atlases
Publish digital atlases of the results, and build augmented-reality tools that let people examine spatial data from inside it rather than on a flat map.
06
Research and teaching capacity
Bring new photogrammetry software and an XR laboratory into the faculty’s courses, and open project topics to master’s and doctoral students.
Key actions
The project runs as eight work packages. Work package 1 covers project management and reporting; the seven below are the research work.
Work package 2
Urban heat islands
Sentinel-2 and Landsat imagery of Zagreb and Split is collected and pre-processed, then classified to compute vegetation index, land surface temperature and environmental criticality index values for May to September each year. Each project year closes with a report on what that season’s data showed.
Work package 3
Water quality assessment
The same satellite systems are used over Lake Vrana near Biograd na Moru throughout the year, estimating optical water-quality parameters from combinations of spectral bands. The resulting annual series shows how concentrations shift through the seasons and from one year to the next.
Work package 4
Landslides and erosion
Field campaigns at Šandrovec and Kutina, twice a year, collect photogrammetric data that is then processed using computer vision and machine learning. The work is carried out together with the Croatian Geological Institute so that the methods reach practitioners, and it doubles as teaching material in photogrammetry courses.
Work package 5
Visualisation and atlases
Results are presented both as conventional cartographic products and through augmented and virtual reality, built in a new XR laboratory at the faculty. In parallel, the vegetation, temperature and criticality results are compiled into digital atlases — one per project year — that stay available to other researchers afterwards.
Work package 6
Analysis and data relationships
Everything the project collects goes through multi-criteria and GIS analysis to establish which data types matter for which kind of decision, and how they relate to each other. Those relationships are what determine how the data should be structured before it is ever stored.
Work package 7
The data warehouse concept
Drawing on those analyses, the project sets out a structure for a future geospatial data lake and warehouse, using its own four datasets as the worked example. The concept is designed to align with Croatia’s National Spatial Data Infrastructure, so that later results can connect to it.
Work package 8
Dissemination
The team publishes in peer-reviewed journals and presents at international conferences throughout the four years, and maintains this website. Master’s and doctoral students take on project topics, so the results reach the next group of researchers directly rather than only through publications.
Partners
University of Zagreb, Faculty of Geodesy
Host institution. Leads the remote sensing, photogrammetry and cartography work.
Andrija Krtalić — principal investigator
Dubravko Gajski — landslides and erosion
Vesna Poslončec Petrić — visualisation and atlases
University of Split, FESB
Multi-criteria analysis and machine learning across the project’s combined datasets.
Ljiljana Šerić
University of Split, FGAG
Thermal behaviour of construction materials, and heat island analysis for the Split study area.
University of Zadar, Centre for Geospatial Technologies
Monitoring of standing water bodies, contributing to the Lake Vrana work package.
University North
Department of Geodesy and Geoinformatics. Geospatial analysis and digital terrain modelling.
State Geodetic Administration
Advises on how the project’s data structure can align with Croatia’s National Spatial Data Infrastructure.
Croatian Geological Institute
Collaborates on landslide detection and monitoring at the Šandrovec and Kutina sites.
About the organisations
University of Zagreb, Faculty of Geodesy
The faculty teaches and researches geodesy and geoinformatics — surveying, photogrammetry, remote sensing, cartography and geographic information systems. It hosts GEOSKLAD and runs the project’s satellite image processing, photogrammetric survey and cartographic work, and will house the XR laboratory established during the project.
University of Split — FESB and FGAG
Two faculties of the University of Split contribute complementary expertise. FESB works on artificial intelligence and distributed systems applied to environmental management, and leads the machine learning and multi-criteria analysis. FGAG brings civil engineering, architecture and geodesy, and covers the thermal behaviour of construction materials and the Split study area.
Partner institutes and public bodies
The University of Zadar’s Centre for Geospatial Technologies contributes experience in satellite monitoring of inland waters. University North’s Department of Geodesy and Geoinformatics contributes geospatial analysis and terrain modelling. The State Geodetic Administration, which maintains Croatia’s National Spatial Data Infrastructure, advises on how project outputs can connect to it, and the Croatian Geological Institute collaborates on landslide monitoring in the field.
Co-funded by the European Union
Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or the granting authority. Neither the European Union nor the granting authority can be held responsible for them.
GEOSKLAD — Concept for a geospatial data warehouse to support environmental decision-making. Carried out at the University of Zagreb, Faculty of Geodesy, under its call for institutional research projects. November 2025 – September 2029.