Description of the study area
The Yayu afro-montane forest is found in the Illubabor Zone, southwest of the country at about 550 Km from the capital, Addis Ababa. The geographic location is between 8⁰4′ 56.05″− 8⁰ 24′ 40.46″ N latitude and 35⁰ 44′53.85″− 36⁰ 5′12.23″ E longitudes (Fig. 1). Large part of the Yanu afro-montane forest is protected as a Forest Biosphere Reserve. The forest is part of the last remaining intact patches of natural forests in the southwest region. The forest has multiple economic, social and environmental benefits. It provides non-timber forest products, mainly spices, honey, and herbal medicine to rural communities for their livelihoods. The forest contains one of the largest forest biomass in the country and hence significantly contributes to climate change mitigation. Besides, the Yayu forest is one of the last remaining montane-rainforests containing wild Coffee arabica gene pool populations in Ethiopia. The forest site is effectively serving as an in situ conservation forest for the wild Coffee arabica population gene pool (Gole et al. 2008; Schuit et al. 2021). Coffee makes the largest share of living for the local communities. The climate is characterized by hot and humid tropical climate with a mean annual temperature of 25°C, varying between 12.7°C and 26.1°C. The region receives high mean annual rainfall of about 2100 mm, with high annual variability ranging from 1400 to 3000 (Gole et al. 2008).
The topography is complex with undulating hills and valleys dissected by several small streams draining into the Geba and Dogi Rivers. The elevation ranges between 1217 m.a.s.l at the valley bottom to 2583 m.a.s.l at the highest point in the watershed (Fig. 2). The valley gorges and the mountains are steep slopes and not easily accessible. As result, the dense and large patches of the forests are located in these parts of the landscape.
The land use land cover was mapped from a Landsat 8 dry season imagery of 2018. The forest land constitutes the largest cover with about 62 % followed by cultivated agricultural land constituting about 30 % of the total cover. The rest of the landscape is covered with shrub lands (3 %), settlements (2.7 %) and wetlands (2.3 %) (Fig. 3). Although the forest area is registered as a National Forest Priority Area and a Biosphere reserve, the local communities are highly dependent on the forest mainly for harvesting natural coffee, spices and honey production. Thus, the Biosphere reserve forest has three functional zones allowing farmers to harvest non-timber forest products in the transition and buffer zones while leaving the core zone as access-restricted conservation zone (located primarily in the valleys and mountains). As shown in Figure 3 below, the dark green covers are the dense forests designated as core zones in the inaccessible high altitude steep mountain and in the low altitude river valleys in the Yayu forest. The landscape in the middle altitude landscape are the buffer and transitions zones, where agricultural cultivation is practices with strict management (Gole et al. 2008)
Three data sources were used for the study. The Landsat 8 image, dated February 2018, was used to classify the land use land cover map and extract the forest cover area of the Yayu forest biosphere reserve. Vegetation parameter data for biomass estimation were directly measured in the field using vegetation sampling plots. Vegetation indices (VIs), biophysical variables (BPVs) and relevant bands were derived from Sentinel-2 imagery (Fig. 4).
The Sentinel-2 satellite imagery, taken in the dry season of February 2018, was downloaded from the open access European Space Agency (ESA) hub. The images were pre-processed using the Sentinel Application Platform (SNAP) and quantum GIS (QGIS). The Sentinel-2 Multispectral instrument (MSI) with swath width of 290 km was Ortho-rectified to UTM Zone 37N projection and a radiometric correction was done to reduce atmospheric and sun angle effects (Baillarin et al. 2012). The image was transformed from radiance to surface reflectance by applying the Dark Object Subtraction (DOS) method using the semi-automatic classification plugin (SCP) in QGIS software. The DOS method removes the darkest pixel in each band that might be affected by atmospheric scattering (Chavez 1988). The blue, green, red and near infrared bands, with 10 m resolution, were resampled into a 20 m resolution using ArcGIS software to correspond with the 20 m vegetation sampling plot size of the field data measurement. The Sentinell-2 MSI was used for deriving multi-spectral bands, vegetation indices (VIs) and biophysical variables (BPVs) (Fig. 4).
Vegetation indices (VI) extraction
The vegetation indices for biomass estimation in this study were extracted from the Sentinel-2 image (Table 1). In a remotely sensed data, a vegetation index is a spectral transformation of two or more bands designed to enhance the contribution of vegetation properties and allow reliable spatial and temporal inter-comparisons of terrestrial photosynthetic activity and canopy structural variations (Huete et al. 2000). Vegetation indices extracted from Satellite data have emerged as important tools in monitoring, mapping and managing terrestrial vegetation as the indices provide radiometric measurement of the quantity, structure and condition of vegetation, and effectively serve as useful indicators of seasonal and inter-annual variations.
There are many VIs with similar functionality and most of them use the inverse relationship between red and near-infrared reflectance associated with healthy green vegetation. The measurements of vegetation attributes include leaf area index (LAI), green leaf area index (GLAI), percent green cover or fractional green cover, chlorophyll content, green biomass and absorbed photosynthetically active radiation (APAR). According to Bannari et al. (1995), VIs are normally classified based on a range of attributes such as the number of spectral bands (2 or greater than 2); the method of calculations (ratio or orthogonal), depending on the required objective; and the historical development (as first generation VIs or second generation VIs). In order to compare the effectiveness of different VIs, Lyon et al. (1998) classified seven types of VIs based on their computational methods (Subtraction, Division or Rational Transform). With the advancement in hyper-spectral remote sensing technology, high-resolution reflectance spectrums are now available to be used along the traditional multispectral VIs. Besides, VIs have also been developed to be specifically used with hyper-spectral data such as the use of Narrow Band Vegetation Indices.
Biophysical variables (BPVs) extraction
Surface biophysical or canopy properties provide an understanding of the physics of the interactions between solar radiation and vegetation elements (Asrar et al. 1989). Surface parameter retrieval from satellite remote sensing data has been one of the major sources to obtain surface parameters because it relates the vegetation characteristics to its spectral signature or reflectance value thereby providing reasonable estimates of vegetation properties across various spectral, spatial and temporal scales (Asrar et al. 1989). According to Widlowski et al. (2004), biophysical variables describe the spatial distribution of vegetation state and dynamics, thus, are useful for biomass estimation. The vegetation indices and biophysical variables were computed using the ArcGIS and SNAP software. The indices were selected based on their performances in biomass estimation in earlier studies (Table 1). The vegetation index map layers were produced using QGIS and ArcMap (Fig. 5 and 6).
Vegetation parameter measurement from sampling plots
A total of 20 randomly drawn sample plots were used to measure the AGB biomass samples from the forest. The vegetation parameter (tree parameters) such as Diameter at Breast Height (DBH) and height (H) were measured in a 20 m X 20 m (400m2) sampling plot, which were randomly generated from the forest map using ArcGIS. The sampling plot coordinates were used as references to locate the plots on the ground and within the transitional, buffer and core zones of the Yayu Biosphere reserve forest. Within each plot, all trees with ≥5 cm diameter and H of > 1.3 m were recorded and measured for DBH and height. The DBH was measured using diameter tape while H was measured using Sunnto clinometer. The field data were used for validating the biomass modeling outputs and to serve as a ground truth data. For most vegetation types in the tropics, a relationship is established for measurable tree parameters and forest stand parameters such as volume and biomass, which are often difficult for a direct measurement (Husch et al. 2003). Hence, already established allometric equations are often used to estimate the biomass by using tree parameter data.
Extraction of the pixel values of predictor variables
The pixel values for each variable derived from the Sentinel-2 image were extracted using zonal statistics in ArcGIS. The field plot geographical location (latitude and longitude) points were used as references to match the pixels as shown in the figure below (Fig. 7). The extracted pixel values for each variable were exported in CSV (comma separated variable) data formats.
Above ground biomass and carbon stock estimation
The above ground biomass and carbon stock were quantified using an allometric equation with input data from the tree parameter measurements such as DBH and H in the field. Besides, specific wood density, which is the dry mass of a unit volume of fresh wood of trees, is used to convert the wood volume into biomass and carbon estimate. The allometric equation selected for this study was established for tropical forest biomass estimation and has been widely applied in similar studies (Chave et al. 2014). The selected equation was applicable for the Yayu forest because of the climatic conditions (mainly rainfall), which is a key parameter for allometric equation determination and vegetation biomass development. The wood density values were species specific and obtained from a secondary source (Gisel et al. 1992).
AGB = 0.0673 x (ρD2H) 0.976 Equ. 1
Where, AGB is Above-ground biomass (g), ρ is specific wood density (g/cm3), D2 is diameter at breast height (DBH) (cm); H is height of tree (m). The above-ground biomass was converted into carbon equivalent using the biomass conversion factor or carbon fraction of 0.47 IPCC (2006).
C = AGB x CF Equ. 2
Where, C is Carbon stock (g), and CF is Carbon Fraction of above ground biomass
Data analysis (Correlation, regression analysis and model development)
The forest biomass data measured from the field and the extracted variables from the Sentinel-2 images were organized into a spreadsheet with CSV format. Correlation between the biomass estimates from the field and variables from the Sentinel images were tested using SPSS software. Those variables having significant correlation with the measured biomass data were identified, selected and a regression analysis was performed between the measured biomass and the vegetation indices to develop a biomass prediction model.
The model was then evaluated based on the magnitude of the Root Mean Square Error (RMSE) and coefficient of determination (r2). The best model was developed by integrating those variables with high r2 and a low RMSE. The equation obtained from the regression model was then used to estimate AGB. The r2 was preferred since it has a standard measure with values ranging from 0 to 1. The r2 also shows the percentage of the variability explained by the model (Husch et al. 2003). This makes it easy to understand the relationship between the independent (indices) and dependent variable (biomass) (Peters 2007). The significance of the model was assessed using the P-Value at α = 0.05. For those significant indices, the equation obtained from the regression model was then used to estimate AGB.