Past and future prediction of land cover land use change based on earth observation data by the CA–Markov model: a case study from Duhok governorate, Iraq

Abstract

Understanding land use land cover change (LULCC) dynamics is crucial for sustaining the integrity of structure and function of ecosystems. As such, frequent measuring and monitoring of LULCC are necessary. Over the last four decades, Duhok governorate in the north of Iraq has undergone sweeping changes caused mainly by anthropogenic factors (e.g. population growth). This study used geospatial techniques and the synergy Cellular Automata (CA)–Markov approach to quantify past, current and model the future changes of LULC. The maximum likelihood classifier (MLC) was employed to conduct classification for three consecutive-year Landsat imagery (i.e. 1988, 2008 and 2017). From the classified imageries, three LULC maps with several classes were created and then, change detection analysis was implied. The classified (1988–2008) and (2008–2017) LULC maps were incorporated into the hybrid model to predict LULC maps for 2017 and 2060, respectively. The classified 2017 LULC maps were used as a reference to validate the model output for 2017. Relatively high accuracy agreements were achieved between the classified and the modelled maps (Kno= 0.8315, Klocation= 0.8267, Kstandard = 0.7978). The model classes estimated for 2060 compared to the classified 2017 LULC classes revealed that dense forest, sparse forest, agricultural land and barren area would decrease by −26.26% (from 327.08 to 241.08 km ²), −0.76% (from 2372.29 to 2355.82 km ²), −5.86% (from 973.21 to 916.27 km ²) and −10.03% (from 2918.9–2626.19 km ²), respectively. In contrast, the urban area would significantly increase by 271.19%, (from 161.99 to 602.19.8 km ²). Dense forest in Duhok governorate has seen remarkable decline from 1988 to 2017, and future predictions demonstrated that the declining trend would continue. Dense forest would predominantly convert to sparse forest and barren areas, suggesting forest thinning and clearing. Urban areas were the most dynamic cover types that increased significantly between 1998 and 2017. This trend would continue to increase from 2.36% (2017) to 8.76% (2060). Urbanization would be predominantly at the cost of agricultural land and barren area. Information on spatiotemporal dynamics of LULCC has been proved as an effective measure for maintaining the integrity of the ecosystem components through sustainable planning and management actions.

More information

Identifiers

Catalogue record

Export record

ASCII CitationAtomBibTeXData Cite XMLDublin CoreDublin CoreEP3 XMLEndNoteHTML CitationHTML CitationHTML ListJSONMETSMODSMPEG-21 DIDLOpenURL ContextObjectOpenURL ContextObject in SpanRDF+N-TriplesRDF+N3RDF+XMLRIOXX2 XMLReferReference ManagerSimple Metadata

Altmetrics