Remote Sensing for Biocultural Heritage Preservation in an African Semi-Arid Region: A Case Study of Indigenous Wells in Northern Kenya and Southern Ethiopia

Abstract

The region of Southern Ethiopia (Borana) and Northern Kenya (Marsabit) is characterised by erratic rainfall, limited surface water, aridity, and frequent droughts. An important adaptive response to these conditions, of uncertain antiquity, has been the hand-excavation of a sequence of deep wells at key locations often along seasonal riverbeds and valley bottoms where subterranean aquifers can be tapped. Sophisticated indigenous water management systems have developed to ensure equitable access to these critical water resources, and these are part of well-defined customary institutional leadership structures that govern the community giving rise to a distinctive form of biocultural heritage. These systems, and the wells themselves, are increasingly under threat, however, from climate change, demographic growth, and socio-economic development. To contribute to an assessment of the scale, distribution and intensity of these threats, this study aimed to evaluate the land-use land-cover (LULC) and precipitation changes in this semi-arid to arid landscape and their association with, and impact on, the preservation of traditional wells. Multitemporal Landsat 5, 7 and 8 satellite imagery covering the period 1990 to 2020, analysed at a temporal resolution of 10 years, was classified using supervised classification via the Random Forest machine learning method to extract the following classes: bare land, grassland, shrub land, open forest, closed forest, croplands, settlement and waterbodies. Change detection was then applied to identify and quantify changes through time and landscape degradation indices were generated using the Shannon Diversity Index fragmentation index within a 15 km buffer of each well cluster. The results indicated that land cover change was mostly driven by increasing anthropogenic changes with resultant reduction in natural land cover classes. Furthermore, increased fragmentation has occurred within most of the selected buffer distances of the well clusters. The main drivers of change that have directly or indirectly impacted land degradation and the preservation of indigenous water management systems were identified through an analysis of land cover changes in the last 30 years, supporting insights from previous focused group discussions with communities in Kenya and Ethiopia. Our approach showed that remote sensing methods can be used for the spatially explicit mapping of landscape structure around the wells, and ultimately towards assessment of the preservation status of the indigenous wells.

Urban expansion and archaeological sites of Wat Chulamani, the oldest temple in Muang Phitsanulok, Thailand

“…The error matrix and Kappa statistics were then used to assess the accuracy. To accomplish this task, training samples of 5 different classes were taken with the reference of Google Earth, which is a very important tool for mapping of different types of land use/land cover and for accuracy assessment [8][11] [15]. It is free and easy to use in data collection, exploration, and visualization.…”

“…Land use changes are an important aspect of land management [6]. Remote sensing is also increasingly used to map and monitor the environmental setting of archaeological sites [7] and thus evaluate the levels of risk arising from such factors as the indirect consequences of urban expansion [8]. This study aims to apply Geoinformation technology together with Unmanned Aerial Vehicle Survey (UAV) on the analysis of land use changes of archaeological sites of Wat Chulamani and surrounding areas.…”

Changes to water management and declining pastoral resilience in Marsabit County, northern Kenya: The example of Gabra wells

“…Moreover, ongoing research with different pastoralist communities has helped locate numerous additional cultural heritage sites and historic water sources in Marsabit County 1 and to relate these to land use and water management by integrating archeological records with ethnographic information and oral histories (M'Mbogori et al, 2022). Trends in changes to land use through analysis of remote sensing data since 1980 have also been analyzed (Ochungo et al, 2021).…”

“…Apart from the resource management interferences, to mitigate against the negative effects of droughts which have become more frequent in recent years, the government and other external actors are encouraging sedentary lifestyles, distribution of relief food in shopping centers, and selective restocking after loss. Thus, in recent times, urban areas have grown in both areal extent and population density leading to land use and land cover fragmentation around many well clusters in both northern Kenya and southern Ethiopia (although not universally so; Ochungo et al, 2021). Some consequences are that some communal areas that were formerly reserved for pasture during extreme climatic conditions, have now become permanently inhabited, and have lost their traditional function.…”

Mapping Spatiotemporal Changes of Evergreen Forest Patches that are Heritage Sites in Southern Mozambique using Google Earth Engine

“…Agapiou (2017) used satellite images and Google Earth Engine (GEE) for the monitoring and management of cultural heritage sites. Ochungo et al (2022) applied satellite images in GEE to map traditional wells and to describe water management systems. However, not all forms of cultural heritage are readily interpretable in satellite images or in the land-cover patterns.…”