Updating and backdating analyses for mitigating uncertainties in land change modeling: a case study of the Ci Kapundung upper water catchment area, Java Island, Indonesia

Abstract

In developing countries, data gaps are common and lead to uncertainties in land cover change analysis. This study demonstrates how to mitigate uncertainties in modeling land change in the Ci Kapundung upper water catchment area by comparing the outcomes of two simulation phases. A conventional cellular automata (CA)–Markov model was complemented with a multilayer perceptron (MLP) to project future land cover maps in the study area. The CA–Markov–MLP model results in high uncertainties in forested sites where a data gap is apparent in the input data (land cover maps and driver variables) and parameters. The results show that the model accuracy is improved from 47.90% in the first phase to 81.36% in the second phase. Both first and second phases integrate six demographic–economic and environmental drivers in the modeling, but the second phase also incorporates an updating and backdating analysis to revise the base-maps. This study suggests that uncertainties can be mitigated by linking such base-map revision process with the updating and backdating analyses using remote sensing datasets retrieved at different times.

Keywords:

• Land cover change
• cellular automata
• CA–Markov
• multilayer perceptron
• uncertainty

Acknowledgments

The authors thank all reviewers and editors for their constructive comments and suggestions. This research is part of a PhD study funded by The Ministry of Research, Technology and Higher Education of Indonesia, and supported by Institut Teknologi Bandung, Indonesia.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data and codes availability statement

The data and codes that support the findings of this study are available with the identifier(s) at https://doi.org/10.6084/m9.figshare.12277427. The SPOT 6 imagery purchased from Airbus cannot be publicly shared. Sample imagery with the same spectral combination and spatial resolution was used for the demonstration. Other datasets used in this research are included in the link and are available in the public domain of the USGS, Indonesian Geospatial Agency and WorldPop at https://earthexplorer.usgs.gov/, http://tides.big.go.id/DEMNAS/, and https://www.worldpop.org/doi/10.5258/SOTON/WP00114, respectively.

Additional information

Funding

This work was supported by The Ministry of Research, Technology and Higher Education of Indonesia.

Notes on contributors

Medria Shekar Rani

Medria Shekar Rani is a lecturer at the Architectural Design Research Group within the School of Architecture, Planning, and Policy Development at Institut Teknologi Bandung, Indonesia. She holds a PhD in Landscape Architecture from the University of Sheffield. Her research focuses on assessing the impact of landscape change on ecosystem services using models, including the factors influencing model accuracy. Medria conducted the remote sensing and land change modeling analyses, as well as writing this manuscript.

Ross Cameron

Ross Cameron is Director of Research, Department of Landscape Architecture, University of Sheffield. He studies the ecosystem services provided by different types of vegetation, including potential to mitigate flooding through rainwater capture, infiltration and dissipation back to the atmosphere. This research is applied at different scales varying from the action of an individual root, up to whole landscape/regional scales. Ross contributes to this paper by advising on how vegetation is depicted and integrated with other land use typologies.

Olaf Schroth

Olaf Schroth is Germanýs first professor in geodesign and teaches at the Faculty of Landscape Architecture, Hochschule Weihenstephan-Triesdorf (HSWT) near Munich in the undergraduate landscape architecture and landscape construction programs, the International Master of Landscape Architecture IMLA and the new Master in Climate Change Management. Since 2021, Olaf has also a research position as part of the Bavarian High Tech Agenda HTA. His research is addressing geodesign, digital tools in landscape planning, landscape visualization and building information models. Olaf supervised the initial research design and advised on the collection, classification and analyses of the remote sensing data.

Eckart Lange

Eckart Lange is Professor Emeritus of Landscape at the University of Sheffield. Prior to joining the University of Sheffield in 2004, from 1990 until 2004 he was Senior Researcher in the City and Landscape Network at ETH Zürich. From 2008 to 2016 he served as a member of the scientific committee of the European Environment Agency. He held visiting appointments e.g. at the University of Tokyo and since 2017 is Visiting Professor at Tongji University, Shanghai. Eckart supervised the research and advised on the conceptualization of the research, contributed to the methodology as well as the reviewing and editing of the manuscript.