Monitoring Continental Wetland Dynamics and Drivers of Changes Using Google Earth Engine

Introduction

Wetlands are rich and diverse and host nearly 10% of the world's biodiversity. They play several functions and provide various services to humans (Gardner and Finlayson, 2018). However, wetlands remain the most threatened ecosystems due to climate change and human activities (Dixon et al., 2016) and continue to disappear globally. Authors have investigated relations between surface water and climate variation or agriculture activities (Poudel et al., 2013). In Moroccan context, studies confirmed that the annual rainfall underwent significant downward trends and would certainly have a significant impact on wetlands (Sebbar, 2013). Also, emerging agriculture strategies promoted agriculture intensification even within arid lands. Intensive use of ground water has been subsideized (PMV, 2008). In mountainous lands such as Middle Atlas which was mainly devoted to extensive pastoral practices have been reoriented toward arboriculture with as consequence the development and the sinking of boreholes and pumping in the aquifers.

Area of the surface of water bodies have significant inter-annual and intra-annual variations. Unfortunately, due to resources lack, spatial and temporal information of surface water that is mandatory to guide conservation and management strategies remain scarce and a common challenge to scientist and managers in several regions mainly in low and middle-income countries.

To deal with the absence of an in-situ monitoring of the state of these ecosystems, earth observation offers opportunities for periodic and systematic monitoring of these areas and become an important data source for monitoring water surface dynamics. Open access satellite images, mainly acquired by Landsat, MODIS and Sentinel satellites, have opened up very interesting development perspectives with well-documented case studies for wetlands monitoring (Pekel et al., 2016). Indeed, specific indices were developped and are based on the physical characteristic of water relying on its lower reflection in near and shortwave infrared wavelength of the electormagnetic spectrum (Smith and Baker, 1981). The integration and processing of such data become possible with the recent development of cloud-based geospatial analysis platforms dedicated to spatial Big Data integrating both image archives from different providers, processing algorithms and distributed processing capabilities. As exemple, Google Earth Engine (GEE) is widely used.

Dayet Aoua, as natural lake of the Middle Atlas, is facing challenging issue during the last years with huge variability of its water surface and long periods of total drying up. Our study contribute to detect and to assess spatiotemporal evolution of Dayet Aoua water surface and the Land Use and Land-Cover (LULC) change between 1984 and 2019. Due to the absence of in situ data, authors focused on the use of remote sensing moderate-spatial-resolution images that will be analyzed within a cloud-based computing platform.

Materials And Methods

Study Area

The lake Dayet Aoua is located in the eastern part of the Middle Atlas (northeast of Ifrane) at 1460 m (Figure 1) with an area of 150 ha, a width of 1100 m, a radius of 3500 m and a maximum volume of 2,5 million m³ (Chillasse and Dakki, 2004). It is fed mainly by the river “Oued El Kantra” at times of high flow and occasionally by other water points with relatively lower flow and the waterbody surface varies according the seasons and the years with a maximum depth around 5 meter. The lake is surrounded by natural forests composed mainly of Cedar, holm oak and Pines in addition to collective lands and agricultural lands

Figure 1.

Location of the Lake Dayet Aoua, Morocco

In addition to the body of the lake, a surrounding area with a radius of 6 km to the lake centroid was defined to have a sufficiently broad vision of land occupation in the territory and thus to analyze the various transitions and changes in this land occupation between 1984 and 2019.