**1. Introduction**

Forests form a critical component of global carbon cycle [1] in form of woody above ground biomass. Their destruction through deforestation disrupts this cycle leading to serious negative consequences on ecosystem functioning. Deforestation may be caused by both natural and anthropogenic factors. Anthropogenic factors included population growth, technological advancements and cultural norms, and these may accelerate deforestation at both local and regional levels [2, 3]. This in turn threatens carbon storage, watershed protection and biodiversity particularly in developing countries [4, 5] where they have been rampant. Forest cover is the most important indicator of forest degradation. The 2015 forest resource assessment (FRA) report of the Food and Agricultural Organization (FAO) indicated that the global forest cover was about 4 billion hectares [6] (about 30.6% of world's land area). Tropical forests covered about 20% of the global land area which dropped to less than 7% at the end of the 20th century [1]. The tropical region experienced the greatest loss and acquisition of forests among the four climate domains of the world, and the highest loss ratio (3.6 to >50% tree cover), indicating the prevalence of the dynamics of deforestation [4].

Conservation of forests improves livelihoods of communities dependent on them for various goods and services. Globally, over 1.6 billion people rely on forests as a source of food, fuel, medicines, water and for cultural use either directly or indirectly [7]. All over the world, forests play an important role in minimizing greenhouse gas emissions [8] and removing excess carbon dioxide from the atmosphere through carbon sequestration. In sub-Saharan Africa, at least 80% of the urban population and 90% rural population depend on charcoal as a source of energy for cooking and warming houses [9, 10]. The main source of energy for cooking in Somalia is charcoal [11–13] which is obtained from forests through deforestation. Furthermore, Somalia being a pastoralist country depends on trees and shrubs for livestock feeds during dry seasons [14, 15]. The resultant forest destruction could lead to devastating impacts on the natural resources which are highly depended on by the pastoral communities [16]. Foreign demand for charcoal has also escalated deforestation rates in Somalia with 4.4 million trees cut down annually for making of export charcoal [13]. As a major export product, charcoal is the major source of income for 70% of the poor and middle-income pastoralists since it requires small capital for production [17–19]. This has significantly contributed to destruction of the country's natural forest resources [20].

In addition to the basic conservation concern about deforestation in Somalia, the international "Reducing Emissions from Deforestation and forest Degradation" (REDD+) initiative has recently become a major component of forest conservation in the country. This is a global effort aimed at conserving, sustainably managing, enhancing, and monitoring forests [21] and thus requires the government through the forestry department to account for the current forest extents for sustainable forest management. This may further require quantification of greenhouse gas emissions and reductions from forests. These therefore necessities quantification of forest cover changes across the country. However, quantifying forest cover changes through field surveys is limited by time and resource availability.

Remotely sensed data however provides alternative means for monitoring forest cover changes [22, 23]. Temporal and spatial components of observational remotely sensed data have improved the quality and the quantity of satellite data by enhancing their applicability in monitoring environmental changes. A number of satellite-derived products have been developed with a focus on global forest cover. MODIS images that have been captured by Terra and Aqua satellites since 2000 have enabled formation of the first annual forest cover product, MODIS Vegetation Continuous Field (VCF) [24, 25]. This product is limited by the patchiness of tree cover changes which have smaller spatial sizes compared to the MODIS VCF [26]. This deficiency led to the production of Landsat based global continuous field cover product (30 m) [27]. The Landsat based product has finer spatial resolution compared to the MODIS based VCF and thus facilitates a more accurate forest cover change assessment. Besides the finer spatial resolution, Landsat data also have high amounts of cloud cover contamination as well as infrequent revisit coverage [28]. Due to this limitation global mosaics were produced for the years 1975, 1990, 2000, 2005 and 2010 [29]. However, more recently a global 30 m forest cover change product was published during 2000–2012 [4]. A latest version of this product is available for the period between 2000 and 2019.

*Remote Sensing Based Quantification of Forest Cover Change in Somalia for the Period… DOI: http://dx.doi.org/10.5772/intechopen.99365*

Even with the increased availability of high-resolution satellite data, it has always been difficult analyzing high spatial resolution satellite data, for example, at a country level. This is because at such a large scale, huge volumes of data are required to cover the entire country [24]. Google Earth Engine (GEE), a cloudbased platform however overcomes this challenge since it allows for processing of huge geospatial datasets [29]. Satellite resources have been used to monitor forest cover changes in various countries across the globe [4]. In the case of Somalia, there have been a few applications but none have quantified forest cover change across the entire country. In this study we utilize the historical earth observation datasets to map and quantify forest cover over time in Somalia by utilizing the capabilities of Google Earth Engine (GEE). The main objective was to evaluate the spatial and temporal patterns of forest cover from 2000 to 2019 across Somalia.
