Assessing spatio-temporal landscape changes in the Ethiopian Highlands with Photogrammetry, Remote Sensing, and Geographic Information Systems
Afromontane regions are important biodiversity hotspots that provide essential ecosystem services of global significance. These regions have experienced severe landscape changes during the last half century. The Bale Mountains ecosystem is one of such regions that is increasingly threatened by land...
Saved in:
| Main Author: | |
|---|---|
| Contributors: | |
| Format: | Doctoral Thesis |
| Language: | English |
| Published: |
Philipps-Universität Marburg
2024
|
| Subjects: | |
| Online Access: | PDF Full Text |
| Tags: |
No Tags, Be the first to tag this record!
|
| Summary: | Afromontane regions are important biodiversity hotspots that provide essential ecosystem services of global significance. These regions have experienced severe landscape changes during the last half century. The Bale Mountains ecosystem is one of such regions that is increasingly threatened by land use and anthropogenic human interference. In addition, the landscape is also changing due to a biotic fossorial rodent mammal endemic to the Bale Mountains in Ethiopia and ecosystem engineer; the giant root rat (GRR; Tachyorycytes macrocephalus). Yet, detailed evidence on the spatio-temporal changes of the Bale Mountains landscape over the last half century is lacking. Consequently, the historical anthropogenic and biogenic contribution to the landscape change remains uncertain. Therefore my PhD study focuses on assessing the spatio-temporal changes of the Bale Mountains landscape using photogrammetry, remote sensing, geographic information systems, and deep learning. More specifically, the PhD study aims to (1) reconstruct the landscape for the last half century at larger spatial extent and very high spatial resolution, (2) evaluate Ericaceous cover change and its impact on the mountain ecosystem, (3) quantify landscape volumetric change induced by human activities (red ash quarrying), and (4) quantify and map the spatio-temporal changes of mounds created by the GRR. The study also provides further insights about the major driving factors for the landscape change.
In Manuscript 1, the first digital elevation models (DEMs) and orthomosaics for the years 1967 and 1984, at very high spatial resolution (0.84 and 0.98 m, respectively), were generated at larger spatial extent (5730 km2) using historical aerial photographs and Structure-from-Motion Multi-View Stereo Photogrammetry in the Bale Mountains. The accuracy of the generated DEMs was assessed using field ground control points. A comparison was also made between the reconstructed DEMs and five readily available open source DEMs. Our results showed that the generated DEMs have the best accuracy and spatial resolution. Our datasets will help the scientific community in addressing questions related to environmental changes in the Bale Mountains and afro-alpine ecosystems.
In Manuscript 2, 1967 and 1984 orthomosaics, and 2017 SPOT satellite imagery were used to map and assess Ericaceous vegetation (Erica trimera, Erica arborea and Erica teniplusia) change and driving factors for this change. Object-based image analysis was used to map the Ericaceous coverage for the three periods. The result showed that the Ericaceous vegetation cover has decreased by 6, 66, and 72 km2 from 1967 to 1984s, 1984 to 2017’s, and from 1967 to 2017’s, respectively. The major factors for the losses are settlement expansion, recurrent fire, uncontrolled agricultural land use practices, grazing pressure, and construction of infrastructure. In conclusion, concerted efforts are needed to mitigate the threats posed by human activities on the Ericaceous ecosystem.
In manuscript 3, the 1967 and 1984 historical aerial photographs, and current in-situ data were used. Structure-from-Motion Multi-View Stereo Photogrammetry, spatial interpolation, and geomorphic change detection analysis were used to generate the DEMs and quantify the volumetric change of two exemplary quarry sites. The result showed that the total net surface volume of the selected sites has decreased by 872,244 m3 over the last half century. This is mainly because of the demand for low-cost material, the unavailability of alternative materials in the region, and lax law enforcement. These landscape changes resulted in habitat loss in the mountain ecosystem. Our results provide valuable insights in mitigating the negative impacts of excavation on the Afroalpine ecosystem.
Manuscript 4 focuses on understanding the spatial distribution, density, and temporal changes of GRR mounds. To detect, count, and map GRR mounds, the 1967 orthomosaic and the 2011 Worldview-2 satellite image data, together with deep learning were employed. The result showed that > 130,000 GRR mounds were formed between the year 1967 and 2011. The overall detection accuracy was 80% and 84% for the year 1967 and 2011, respectively. In addition, the result showed that biogenic factors such as GRRs are shaping the landscape in the Afro-alpine ecosystem of Bale Mountains. The increment of GRR mounds has both positive and negative environmental implications, affecting different aspects of the ecosystem.
This thesis has comprehensively assessed the spatio-temporal landscape changes in the Ethiopian highlands. The study has provided new insights into the extent and nature of changes in this critical Afromontane ecosystem. The degradation of Ericaceous vegetation, significant volumetric changes due to quarrying activities, and the proliferation of GRR mounds have been identified as major contributors to landscape alteration in Bale Mountains. The findings underscore the urgent need for conservation efforts to mitigate the impacts of anthropogenic and biotic factors on this biodiversity hotspot. |
|---|---|
| DOI: | 10.17192/z2025.0088 |