In this research, the impacts of past (1984–2020) and future (2040–2060) changes in Ogun River Basin's land use land cover (LULC) dynamics were studied. LULC changes were classified using the Random Forest algorithm in Google Earth Engine on Landsat imagery from 1984, 2000, and 2020 epochs. The multi-layer perceptron-neural network and Markov Chain in the Land Change Modeler were used to predict the LULC maps under various future scenarios. The results revealed significant alterations in the LULC pattern. Forests, wetlands, and water bodies decreased between the epochs 1984 and 2020 by −29.0%, −58.3%, and −84.5%, respectively, whereas farmland and built-up areas expanded by 104 and 20%, respectively. The Business-as-Usual scenario is anticipated to result in an increase in farmland of 3.73% in 2040 and 3.89% in 2060, respectively, at the detriment of forest cover and wetland areas, which are projected to decrease. Under the afforestation scenario, the forest is predicted to expand at the expense of farmland and built-up areas, which are anticipated to expand by 6% (2040) and 8% (2060), respectively. These results, in combination with those derived from the Markov model, provide the necessary evidence base to support land use planning and the future-proofing of sustainable water resources management strategies in the basin.

  • Anthropogenic induced activities affected the land use land cover changes in the Ogun River Basin.

  • Forested area declined from 1984 to 2020 with increasing farmland and built-up areas.

  • Agricultural land encroachment to forest and wetlands was the major factor that affected the land use changes and the river basin.

The rising population growth, declining economic growth, and physical attributes such as slope, landscapes, soil types, and climate are the primary causes of changes in land use land cover (LULC) changes (Lambin et al. 2003; Setegn et al. 2009; Yalew et al. 2016). Human activities have been identified as one of the major drivers for LULC changes and simultaneous changes in natural ecosystems (WoldeYohannes et al. 2018). Global crisis is currently prioritized by the growing urbanization rates in the majority of developing nations and the implications of this mutation for future environmental and urban planning processes are important (Rawat & Kumar 2015). On a global scale, human-induced environmental changes have made a huge impact on the planet, from the regional to the global level (Rindfuss et al. 2004), especially anthropogenic land cover changes which have jeopardized the ecological system sustainability (Verburg et al. 2011). Urbanization, the development of agricultural land, deforestation, and desertification (Radwan 2019; Liu et al. 2020a, 2020b) are some major changes in land cover that negatively affect both environmental conditions and human activities (Nzunda & Midtgaard 2019; Sloan et al. 2019).

In Africa, Kouassi et al. (2021) reported conversions of forestland to agricultural land by 1.44% and dense forest to degraded forest by 3.44% from 1987 to 2015 in Cote d'Ivore. Mathewos et al. (2022) also predicted that in Ethiopia, cultivated and settlement areas would rise by 6.4 and 6.5%, respectively, while grassland and forest areas would decline by 22.3 and 63.8%, respectively, by 2050. Several studies on the impacts of LULC changes on urban expansion in Nigeria have shown an abrupt rise in this growth; for instance, Nzoiwu et al. (2017) reported that the built-up area expanded from 9.6 km2 in 1986 to 21.5 km2 in 2015 at the expense of natural vegetation in South-East Nigeria. Furthermore, Fashae et al. (2020) found similar results of an increased urban cover from 341.7 to 520.6 km2 in 1984–2019 in Southwestern Nigeria. This rapid LULC change is a severe impediment to sustainable growth since it adversely affects forest vegetation, risk of flooding, urban planning, agriculture, and the availability of water resources (Dimobe et al. 2017; Akinyemi 2021; Nut et al. 2021).

The Ogun River Basin (ORB), which is in the southwestern region of Nigeria, is experiencing increasing urbanization as a result of the impact of changing land use along the periphery of significant urban concentrations (Awoniran et al. 2014). The primary threat that the basin is facing is the accelerating rate of anthropogenic activities including peri-urbanization, increased population, mainly in the lower parts of the basin, industrialization, land use expansion, and socio-economic activities. According to Tobore et al. (2021), the Odeda peri-urban area of Ogun River experienced an enormous increase in farmland and built-up areas as a result of increased human activities, such as agricultural land use expansion, soil and water pollution, and biodiversity loss. Similarly, various studies in the lower ORB have noted an expansion in the built-up and agricultural areas during the last 20 years (Odunuga & Oyebande 2007; Awoniran et al. 2014; Ojo et al. 2021; Tobore & Bamidele 2022). Studies have not really attempted to ascertain how the past and future LULC dynamics would affect the basin as a whole, which supports the Ogun River's activities with a focus on sustainable water management.

The assembling and processing remote sensing imagery and other ancillary data, with recent big data and cloud computing methodologies (Casu et al. 2017; Huang et al. 2021), have provided new remedies for multitemporal evaluation glitches, which is the open-source JavaScript Application Programming Interface (API) known as Google Earth Engine (GEE), which enables users to process large amounts of data using cloud computing methods and a multi-petabyte catalog of remote sensing data (Gorelick et al. 2017; Xie et al. 2019; Phan et al. 2020; Floreano & de Moraes 2021; Dubertret et al. 2022). There are a multitude of machine learning techniques that are used to detect the LULC (Nery et al. 2016; Dimobe et al. 2017; Hackman et al. 2017) but in comparison with the machine learning algorithms by Talukdar et al. (2020), Random Forest (RF) emerged as the best-performing algorithm.

Nevertheless, several methods have been developed to predict future LULC under the Land Change Modeler (LCM) (Dey et al. 2021, Tobore & Bamidele 2022; Yang et al. 2022; Yangouliba et al. 2022). The multi-layer perceptron-neural network (MLP-NN) and Markov Chain Model is a robust supervised training technique that depends on back propagation (BP) and this model is the most efficient method for this study's prediction (Mishra et al. 2014; Eastman 2020; Hussien et al. 2022). As a result of this finding, the objective of this study is to evaluate the historical (1984–2020) LULC using the RF algorithm and predict the future (2040–2060) LULC changes using the MLP-NN and Markov Chain Model embedded in the LCM due to anthropogenic activities over the ORB.

Study area

The ORB (shown in Figure 1) has a landmass of 23,000 km2 and is located in Nigeria's rain forest zone between latitudes 6° 26′ N and 9° 10′ N and longitudes 2° 28′ E and 4° 8′ E. The Ogun River originates from the Igaran hills at an elevation of around 530 m above mean sea level and flows southward across a distance of 480 km before draining into the Lagos Lagoon (Awe et al. 2020). The river basin occupies 2.58% of Nigeria's landmass and cuts across three states (Lagos, Ogun, and Oyo states), housing two reservoirs (Ikere gorge and Oyan dam) which are heavily populated (Awoniran et al. 2014; Komolafe et al. 2020).
Figure 1

Location and elevation maps of the ORB.

Figure 1

Location and elevation maps of the ORB.

Close modal

There are two distinguishable seasons of rainfall in the basin, with the dry season spanning from November to March and the rainy season lasting from April to October. Due to the basin's location in Nigeria's humid subtropical zone, the mean annual precipitation ranges between 1,200 and 1,700 mm and the mean temperature between 22 and 36 °C (Komolafe et al. 2020). The basin population has grown significantly over the past decades as a result of the basin's increased rates of industrialization and urbanization (Awe et al. 2020). The watershed housed two main vegetative zones which are the high forest vegetation found in the north region and the swamp/mangrove forests that cover the southern coastline and flood plains close to the lagoon (Asinwa et al. 2018).

Data acquisition

In this study, Landsat imagery at 30-m resolution was used for the LULC classification provided by the United States Geological Survey (USGS). A 30-m Digital Elevation Model (DEM) from the Shuttle Radar Topographic Mission (SRTM) was downloaded to delineate the topography, rivers, and road network. Data from the Global Positioning System (GPS) were collected for generating training samples. This study utilized the LULC data from Landsat 5 Thematic Mapper (TM) images from 1984 epoch, Landsat 7 (Enhanced Thematic Mapper Plus (ETM+)) images from 2000 epoch, and Landsat 8 (OLI) images from 2020 epoch (shown in Supplementary material S1). The Landsat imagery of 30-m spatial resolution can be used for change detection in LULC transition (Midekisa et al. 2017). Google Earth's classification algorithm was used to obtain a single cloud-free image and images between the desired temporal ranges. In contrast to Landsat 7 (ETM+) and Landsat 8 (OLI) that utilize the Surface Reflectance (SR) image, Landsat 5 (TM) uses raw scenes imagery collection. The GEE Data Catalogue was used to extract these data, which were then used as inputs for LULC analyses.

In order to determine the five primary LULC classes (water bodies, forests, built-up areas, farmland, and wetland areas), a modified LULC classification method (shown in Supplementary material S2) was adopted (Ojo et al. 2021). Three components of the training samples were sourced from the Google Earth map for the 1984 and 2000 epochs while the 2020 epoch was captured by a field survey using stratified random sampling for the five LULC classes. The classification of 1984, 2000, and 2020 images, respectively, employed 375, 290, and 261 training sample sizes. The training samples were obtained from around 70% of the participants who were randomly selected, whereas the validation samples were obtained from 30% of the participants (shown in Supplementary material S3).

LULC prediction, change analysis, and driver's variables

A generative NN with one or more layers between the input layer and the output layer, which is known as the MLP-NN that depends on the BP algorithm, is the main supervised training method adopted (Falahatkar et al. 2011; Mishra et al. 2014; Eastman 2020) and was used for LULC prediction (LCM embedded in the TerrSet Software).

The land change analysis creates maps and graphs of land change rapidly using two historical (1984 and 2000 images) land cover layers, including gains and losses, net change, persistence, and a description of the factors contributing to each transition. The selection of drivers' variables for LULC changes was based on elements that increase or decrease the activity's suitability (Mishra et al. 2014; Wang & Maduako 2018).

Acuracy assessment, LULC classification and predicton scenarios

The relevance of accurate LULC classification has been the distinction between LULC classification and ground truth data (Ewunetu et al. 2021), whereas the confusion matrix evaluates the remote sensing images classification between LULC classes and validation data (Liu et al. 2020a, 2020b). The LULC classification was exclusively based on supervised RF classification with overfitting and noise power processing (Na et al. 2010), and RF provides higher accuracy than single trees and maximum likelihood (Belgiu & Dragut 2016; Teluguntla et al. 2018). Researchers in West Africa have utilized it effectively (Hackman et al. 2017; Thiam et al. 2022; Yangouliba et al. 2022). The MLP-NN technique was used to map the significant transition potentials between one LULC class and another (Larbi et al. 2019). Five major transitions, larger than 50,000 hectares for potential transitions, were taken into consideration in this study. These transitions are wetland to forest, built-up to farmland, farmland to forest, forest to farmland, and forest to wetland determined from the change analysis.

Creating scenarios of potential future LULC changes is essential for a multitude of research areas and land use change is a somewhat unregulated process compared to industrialization in developing nations like Nigeria because of the complicated land tenure system. In this research, two LULC change scenarios were considered: Business-As-Usual (BAU) and afforestation scenarios (illustrated in Supplementary material S4) for future predictions and the future LULC changes for 2040 and 2060 were estimated using the same drivers and observable changes in LULC from 1984 to 2020. In order to create the afforestation scenario, the probability matrix for natural vegetation (forest) was changed at the detriment of farmland and built-up areas by reducing the likelihood that forest would be converted to farmland and built-up areas. The ‘BAU’ scenario was based on the historical history of LULC transitions, which is the expansion of farmland to the detriment of natural vegetation (forests), as indicated in Table 3.

Model validation

The purpose of the validation method is to compare the quality of the observed map for 2020 with the simulated LULC maps for 2020 in order to evaluate the model's accuracy. The initial period of 1984–2000 was taken into consideration in order to replicate the LULC map for 2020 using the Markov Chain Model for the simulation. This method has been applied in various studies to forecast the future LULC (Larbi et al. 2019; Hasan et al. 2020; Yangouliba et al. 2022). The module VALIDATE, which enables the calculation of the Kappa index, was utilized in the model validation procedure (Pontius 2000).

The Kappa index spans from −1 to 1, with positive values denoting better agreement and negative values denoting disagreement (Chaudhuri & Clarke 2014; Roy et al. 2014). The classification of kappa values was as follows: poor less than 0.40, acceptable to good values between 0.40 and 0.75, and excellent greater than 0.75 (Roy et al. 2014).

These equations were used to calculate the Kappa coefficient (Tadele et al. 2017; Teluguntla et al. 2018):
(1)
where N is the number of observations, n is the total number of pixels (observations), xii represents the number of observations in row i and column i, Xi+ is the marginal total of row i, and X+i is the marginal total of column i.

Data preparation and image classification

Stacking, masking, image scaling, and band selection are all part of image processing. In order to eliminate clouds and replace the region with cloud-free images, masking and filling approaches are utilized. The cloud masking input reduced by <20% of the cloudiness from the Landsat images and a cloud-free composite median was created in order to construct a single imagery collection. RF is a regression technique with an ensemble of trees that were built from a training set of data and internally validated to produce a response prediction providing predictors for future observations (Boulesteix et al. 2012; Kulkarni & Sinha 2013). Other characteristics, such as the normalized difference built-up index (NDBI) and the normalized difference vegetation index (NDVI), were taken into consideration. In this study, the NDVI was used to quantify vegetation and was also helpful in determining vegetation density, while the NDBI provides details on the level of urbanization in the region as well as land cover changes (Hackman et al. 2020; Prasomsup et al. 2020; Alademomi et al. 2022).

These formulas were used in GEE to calculate the NDVI and the NDBI as follows:
where NIR means near-infrared and SWIR means shortwave infrared wavelength.

Accuracy and LULC classification

The overall classification accuracy was 80, 89, and 92%, for 1984, 2000, and 2020, respectively, and the Kappa coefficients were 78, 82, and 90%, respectively. Supplementary material S5–S7 show the statistics of the confusion matrix for the 1984, 2000, and 2020 LULC maps, respectively. The accuracy of the most recent LULC maps was higher, which may be attributed to the improved spatial resolution of the satellite imagery. Figure 2 displays the land use and cover maps for the epochs 1984, 2000, and 2020. Table 1 illustrates the area statistics for various LULC classes over time.
Table 1

Statistics of the LULC maps of 1984, 2000, and 2020

1984
2000
2020
1984–20002000–20201984–2020
LULC classesArea (km2)Change (%)Area (km2)Change (%)Area (km2)Change (%)(%) (%)(%)
Water bodies 600.68 2.67 154.87 0.69 92.97 0.41 −74.22 −39.97 −84.52 
Forest 14,632.88 65.00 11,167.58 49.61 10,388.68 46.15 −23.68 −6.98 −29.00 
Built-up 692.76 3.08 767.62 3.41 834.08 3.70 10.81 8.66 20.39 
Farmland 5,198.49 23.09 9,066.49 40.27 10,617.82 47.16 74.41 17.11 104.25 
Wetland 1,388.02 6.17 1,356.28 6.02 579.28 2.57 −2.29 −57.29 −58.27 
Total 22,512.83 100 22,512.83 100 22,512.83 100    
1984
2000
2020
1984–20002000–20201984–2020
LULC classesArea (km2)Change (%)Area (km2)Change (%)Area (km2)Change (%)(%) (%)(%)
Water bodies 600.68 2.67 154.87 0.69 92.97 0.41 −74.22 −39.97 −84.52 
Forest 14,632.88 65.00 11,167.58 49.61 10,388.68 46.15 −23.68 −6.98 −29.00 
Built-up 692.76 3.08 767.62 3.41 834.08 3.70 10.81 8.66 20.39 
Farmland 5,198.49 23.09 9,066.49 40.27 10,617.82 47.16 74.41 17.11 104.25 
Wetland 1,388.02 6.17 1,356.28 6.02 579.28 2.57 −2.29 −57.29 −58.27 
Total 22,512.83 100 22,512.83 100 22,512.83 100    

Note. The overall changes that occurred in the basin between the years 1984 and 2020 revealed a significant decrease in water bodies, forests, and wetland areas while a significant increase in built-up and farmland areas.

Figure 2

LULC maps of 1984, 2000, and 2020. (Note. The above-mentioned result revealed an increase in farmland and built-up areas between the years 1984 and 2020.)

Figure 2

LULC maps of 1984, 2000, and 2020. (Note. The above-mentioned result revealed an increase in farmland and built-up areas between the years 1984 and 2020.)

Close modal

The LULC changes that occurred between 1984 and 2000 include decrease of water bodies, forest areas, and wetland areas with a significant annual change rate of −74.22, −23.68, and −2.29%, respectively, while built-up areas and farmland increased with a significant annual change of 10.81 and 74.41%.

The LULC changes that occurred between 2000 and 2020 include decrease of water bodies, forest areas, and wetland areas with a significant annual change rate of −39.97, −6.98 and −57.29%, respectively, while built-up areas and farmland areas increased with a significant change of 8.66 and 17.11%.

Change contributors, major drivers of LULC dynamics, and transition modeling potentials

The main contributors and causes of expansion were built-up and farmland areas (illustrated in Figure 3) that affect the decline in forested area by 4,228 km2 and could primarily be attributed to the increase in farmland between 1984 and 2000, whereas the decrease in farmland area by 595 km2 was linked to the expansion in built-up areas. The wetland area was also decreased by 309 km2 between 2000 and 2020; generally, the expansion in built-up areas is attributable to farmland and forested areas (shown in Figure 4).
Figure 3

Net change contributors in built-up and farmland areas (in km2) for 1984–2000, 2000–2020, and 1984–2020. (Note. Net contributors to changes were mainly the built-up and farmland areas that declined the forested area.)

Figure 3

Net change contributors in built-up and farmland areas (in km2) for 1984–2000, 2000–2020, and 1984–2020. (Note. Net contributors to changes were mainly the built-up and farmland areas that declined the forested area.)

Close modal
Figure 4

LULC maps for 2020 (classified and simulated). (Note. The images show the validation of the LULC maps for the recent (2020) LULC changes in the basin.)

Figure 4

LULC maps for 2020 (classified and simulated). (Note. The images show the validation of the LULC maps for the recent (2020) LULC changes in the basin.)

Close modal

The major drivers of LULC are the land selection for agriculture and urban encroachment which were greatly influenced by slope and elevation. There will not be any urban development if there are no access points or water sources; hence, the distance to a road and river also plays a significant spatial influence on how rapidly a city grows. Transitions were modeled using a transition sub-model after the predictor variables were specified, and an MLP-NN produced the transition potential maps with an accuracy of more than 81%. The findings from the study of changes in land use and cover showed that farmland and forests are mainly responsible for the major changes that have significantly occurred in urban areas. The LCM takes into account the following transitions for areas larger than 500 km2: forest-farmland, built-up-farmland, farmland-forest, and wetland-forest. Based on the visible indicators of the urban spatial pattern, the driving force behind all of these transitions was the same and an evidence of likelihood which shows a considerable impact on urban growth (illustrated in Supplementary material S8 and S9).

The transition potential maps (shown in Supplementary material S9) from farmland to forest (0.65), built-up to farmland (0.73), and wetland to forest (0.65) showed higher transition while forest to farmland (0.48) showed lower transition, which implies that the transition that occurs in built-up areas is affected by farmland, and likewise, wetland is affected by forest. As the distance to a road and river plays a major role in urban population, slope and elevation also determine all these transitions.

Spatiotemporal trends of anthropogenic disturbances and model validation

The spatial pattern of LULC changes to farmland and built-up areas between the years 1984 and 2000 and 2000 and 2020, respectively, is illustrated in Supplementary material S10 and S11. The analyzed trend represented negative values, which indicate a reverse temporal development, while the positive values indicate an increasing intensity. It is noticeable that the conversion pattern for other LULC is advancing north-eastward, while the pattern of conversion for farmland and built-up areas was toward the south-east from 1984 to 2000. Farmland and built-up areas had a sporadically significant correlation. The intensity of change for farmland (0.40) was greater than built-up areas (0.021) in the first 16 years (1984–2000), and it decreased in the subsequent 20 years for farmland (0.24) compared to built-up areas (0.029) from 2000 to 2020.

The Kappa variation characteristics (VALIDATE module) of Idrisi software was used to compare the simulated and classified LULC maps of 2020. While Klocality analyses the model's grid cell locations, Kno examines the model's linear function. Although there are some minor variations between the two maps of 2020, the results revealed Kno of 0.82 and Klocality of 0.85, showing the module's capacity to model the predicted LULC (shown in Figure 4). For instance, the simulated map (11,192 km2) is greater than the classified map (10,618 km2) for farmland, with a difference of 260 km2 (shown in Table 2). Environmental factors including alterations in the dynamics of land use and land cover changes may be the cause of the significant variations between both the simulated and classified maps for wetland and water bodies.

Table 2

Area statistics of the LULC maps of 2020 (classified and simulated)

Classified
Simulated
LULC unitsArea (km2)Change (%)Area (km2)Change (%)
Water bodies 92.97 0.41 231.42 1.03 
Forest 10,388.68 46.15 9,140.12 40.60 
Built-up 834.08 3.70 573.87 2.55 
Farmland 10,617.82 47.16 11,191.54 49.71 
Wetland 579.28 2.57 1,375.88 6.11 
Total 22,512.83 100.00 22,512.83 100.00 
Classified
Simulated
LULC unitsArea (km2)Change (%)Area (km2)Change (%)
Water bodies 92.97 0.41 231.42 1.03 
Forest 10,388.68 46.15 9,140.12 40.60 
Built-up 834.08 3.70 573.87 2.55 
Farmland 10,617.82 47.16 11,191.54 49.71 
Wetland 579.28 2.57 1,375.88 6.11 
Total 22,512.83 100.00 22,512.83 100.00 

Note. The overall changes between the validated maps of 2020 and the slight changes might be due to the environmental factors.

The predicted LULC change dynamics

The results of the LULC projections for the various LULC change scenarios (illustrated in Figure 5) are revealed under the BAU scenario (shown in Table 3); farmland might potentially expand from 47.16% in 2020 to 48.92% in 2040 at the loss of natural vegetation (forest). Water bodies, wetland, and built-up areas might all increase by 0.42, 0.19, and 3.72%, respectively. Moreover, farmland has expanded from 47.16% in 2020 to 49% in the BAU scenario for 2060, whereas built-up areas and water bodies in the 2040 scenario remain unchanged and wetlands declined by 0.17%.
Figure 5

Predicted LULC maps for 2040 and 2060 (under BAU and afforestation). (Note. The changes between the scenarios of the LULC prediction maps between 2040 and 2060.)

Figure 5

Predicted LULC maps for 2040 and 2060 (under BAU and afforestation). (Note. The changes between the scenarios of the LULC prediction maps between 2040 and 2060.)

Close modal
Table 3

Projected LULC area statistics for 2040 and 2060 relative to the baseline (2020)

LULC unitsBaseline
2020 km2 (%)
2040
BAU km2 (%)
2060
BAU km2 (%)
2040
Afforestation km2 (%)
2060
Afforestation km2 (%)
% Change 2020-BAU 2040% Change 2020-BAU 2060% Change 2020–2040
Afforestation
% Change 2020–2060
Afforestation
Water bodies 92.97 (0.41) 93.47 (0.42) 93.47 (0.42) 93.47 (0.42) 93.47 (0.42) 0.53 0.54 0.53 0.53 
Forest 10,388.68 (46.15) 10,524.14 (46.75) 10,512.82 (46.70) 11,079.96 (49.22) 11,247.15 (49.96) 1.30 1.19 6.65 8.26 
Built-up 834.08 (3.70) 838.25 (3.72) 838.25 (3.72) 474.27 (2.11) 348.31(1.55) 0.50 0.50 −43.14 −58.24 
Farmland 10,617.82 (47.16) 11,013.34 (48.92) 110,331.13 (49.00) 10,570.21 (46.95) 10,562.76 (46.92) 3.73 3.89 −0.45 −0.52 
Wetland 579.28 (2.57) 43.63 (0.19) 37.15 (0.17) 294.92 (1.31) 261.14 (1.16) −92.47 −93.59 −49.09 −54.92 
Total 22,512.83 22,512.83 22,512.83 22,512.83 22,512.83     
LULC unitsBaseline
2020 km2 (%)
2040
BAU km2 (%)
2060
BAU km2 (%)
2040
Afforestation km2 (%)
2060
Afforestation km2 (%)
% Change 2020-BAU 2040% Change 2020-BAU 2060% Change 2020–2040
Afforestation
% Change 2020–2060
Afforestation
Water bodies 92.97 (0.41) 93.47 (0.42) 93.47 (0.42) 93.47 (0.42) 93.47 (0.42) 0.53 0.54 0.53 0.53 
Forest 10,388.68 (46.15) 10,524.14 (46.75) 10,512.82 (46.70) 11,079.96 (49.22) 11,247.15 (49.96) 1.30 1.19 6.65 8.26 
Built-up 834.08 (3.70) 838.25 (3.72) 838.25 (3.72) 474.27 (2.11) 348.31(1.55) 0.50 0.50 −43.14 −58.24 
Farmland 10,617.82 (47.16) 11,013.34 (48.92) 110,331.13 (49.00) 10,570.21 (46.95) 10,562.76 (46.92) 3.73 3.89 −0.45 −0.52 
Wetland 579.28 (2.57) 43.63 (0.19) 37.15 (0.17) 294.92 (1.31) 261.14 (1.16) −92.47 −93.59 −49.09 −54.92 
Total 22,512.83 22,512.83 22,512.83 22,512.83 22,512.83     

Note. The overall changes revealed a decrease in wetlands (BAU) from 2040 to 2060 while a decrease in built-up areas (afforestation) from 2040 to 2060.

The percentage of forest vegetation could increase from 46.15% in 2020 to 49.22% in 2040 and 49.96% in 2060 under the afforestation scenario for 2040, which anticipates a decrease in farmland expansion through increasing vegetation growth while preserving food security. From 47.16% in 2020 to 46.95% in 2040 and 46.92% in 2060, farmland decreased under afforestation scenarios. On the contrary, built-up areas and wetlands will decrease by 2.11 and 1.31% in 2040, as well as 1.55 and 1.16% in 2060, respectively.

Classification accuracy and LCM validation

In the ORB, the rates of industry, urbanization, socio-economic growth, and infrastructural development have all increased. Changes in LULC, human activities, and a sustainable future environment are all interconnected in complex ways (Rimal et al. 2017). This study evaluates how LULC dynamics have changed over a 36-year period in the ORB using satellite imagery. Due to the research area's variability and the probability that farmland and forest were confused for one another, the LULC classification accuracy found in this study can be attributed to these factors. In addition, classifier, space, and time also affect how accurate an LULC classification is and this may be as a result of variations in the atmosphere, surface, and illuminations (Manandhar et al. 2009; Li et al. 2016).

An MLP threshold of 80% is considered to be an acceptable performance according to Eastman (2020), and the LCM efficiently predicted the probable transition changes between the years 1984 and 2000 with an accuracy of 82%. The structure of the model may be ascribed to the model validations, which is based on the comparison between the classified and the simulated LULC of 2020 obtained from this study. The simulated LULC was found to have overestimated and underestimated certain classes. According to Larbi et al.’s (2019) findings, this modification might be linked to the model's structure, which prevented extrapolating stationary changes from the calibration period (1984–2000) to the validation period (2000–2020). The non-linear relationship between humans and environment, which can be challenging for variables and algorithms to capture, makes it difficult to estimate land changes, according to Perez-Vega et al. (2012).

Historical LULC change dynamics analysis

The historical LULC changes showed that from 1984 to 2020, farmland and built-up areas expanded continuously while forests, water bodies, and wetland decreased. In the past 36 years, deforestation (the expansion of agricultural land) and urbanization have resulted in a loss of more than 50% of forest cover. Oyerinde et al. (2015) reported that anthropogenic activities such as excessive bush burning and deforestation have a negative impact on Nigeria's vegetative cover. The reduction in the area covered by wetlands and forests showed that agriculture and urbanization are the primary drivers influencing how land has been utilized in the study area. Similarly, Awoniran et al. (2014) discovered that the conversion of forests as well as shrubs was an indication of increasing anthropogenic activities and a strong interest in urban agriculture in the basin. Also, Alabi et al. (2021) reported that existing anthropogenic activities like deforestation will continue to put forest cover in peril.

Furthermore, the decrease in wetland results is consistent with the findings of Tobore & Bamidele (2022), who reported that unregulated farming activities and built-up intrusion are the primary factors behind wetland anomalies. The expansion in farmland and built-up areas correlates with the findings of studies conducted in a similar region by Akintuyi et al. (2021), Ashaolu et al. (2019), Balogun & Ishola (2017), and Tobore et al. (2021). The rate of urbanization and increase in population is alarming; urban areas in the lower part of the basin increased by 53%, according to Awoniran et al. (2014), to the detriment of forests, wetlands, and water bodies. The implication was that environmentally sensitive places have been vulnerable to urban development without planned mitigation, leading to negative effects on biodiversity and ecosystem degradation (Mmom & Chukwu-Okeah 2011).

In addition, urban migration has increased in Nigeria's major cities over the last few decades as a result of the country's high population density (Idowu 2013). The findings of the historical LULC dynamics are consistent with those of Akintuyi et al. (2021), Onanuga et al. (2022), and Tobore & Bamidele (2022), who discovered an increase in farmland and built-up areas and a decline in natural vegetation and wetlands in the ORB cities of Ogun, Ijebu, and Oyo.

Impacts of future LULC change dynamics on the water resources management of the ORB

Without the introduction of new economic or environmental regulations, the predicted LULC maps under the BAU scenario were based on historical and present population and economic growth. In this scenario, the BAU predicted that between 2020 and 2040 as well as 2060, there would be a significant decrease in wetlands and expansion in farmland and built-up areas to the detriment of water bodies and forests. Overexploitation of land resources, which provides a means of sustenance and employment for those residing in peri-urban regions, could be the cause of the expansion in farmland that has been observed (Tobore et al. 2021). Also, because of the city's uncontrolled population growth, the expansion of built-up areas may compel people to move from urban areas to peri-urban areas (Adu-Gyamfi 2021). Anthropogenic activities including irrigation, industrialization, water pollution, and agricultural land expansion could be considered the primary causes of these changes (Awoniran et al. 2014).

Deforestation is one of the LULC changes that is known to be the main cause of alterations in hydrological processes such as infiltration, evapotranspiration, groundwater, surface runoff, and rainfall interception (Guzha et al. 2018; Gashaw et al. 2019; Ogato et al. 2021). Rapid urbanization has an impact on river water quality, and the increase of farmland may cause synthetic fertilizers to spill or leak into the soil, contaminating local water sources (Roy et al. 2015; Hua 2017; Daba & You 2022). These present findings align with various other findings which indicate that deforestation is a contributing factor to the expansion of farmland and built-up areas (Mekasha et al. 2020; Amankwah et al. 2021); the alteration in wetlands affects soil formation, and also, declining groundwater levels have an influence on water resources (Ballut-Dajud et al. 2022).

Furthermore, under the afforestation scenario, the ORB's farmland would be mostly covered by 46.95% (2040) and 46.92% (2060), respectively, and there would be an increase in the area covered by forests by 49.22% (2040) and 49.96% (2060). The two-afforestation scenario's results are comparable (shown in Table 3). According to some research, an afforestation scenario would result in less agriculture and more forest regeneration (Larbi et al. 2019; Yangouliba et al. 2022).

In summary, the expanded forest cover (conversion of farmland to vegetation cover) under the afforestation scenario would give rise to evapotranspiration by the tree cover (Oliveira et al. 2018) and a decline in surface runoff and water yield (Zhang et al. 2017). Tree cover improves soil infiltration through rainfall and streamflow, thereby increasing water storage and groundwater recharge (Nugroho et al. 2013). Nonetheless, this study has revealed threats to LULC alterations, provided important insights into LULC formation in the future, and also identified areas for future protection, mitigation, and sustainability.

The ORB has undergone significant LULC changes due to induced socio-economic activity during the past three decades. This study analyses the historical (1984–2020) and projected (2040–2060) changes in LULC in the ORB. RF classification was the technique used in GEE, which produced maps for the years 1984, 2000, and 2020. The MLP-NN and Markov Chain predictor of LCM were used to estimate the future LULC (2040 and 2060), and the LULC map of 2020 was selected as a reference year to validate the model.

The research concluded that the LULC in the ORB has changed to the point where it has impacted the spatial pattern and different land uses both inside and outside of the basin. Also, the study observed that human activities such as deforestation, peri-urban migration, and agricultural practices have caused a rising trend in farmland and built-up areas as well as a decrease in forest cover and wetlands. This study's limitations are primarily caused by the cloud cover that degrades the quality of the Landsat images during the rainy season.

Furthermore, the future scenario under BAU projected a steady expansion in farmland and built-up areas at the loss of forest cover and wetland areas in 2040 and 2060. However, the afforestation scenario projected an increase in the forest cover, which would lead to an increase in evapotranspiration by tree cover and reduced surface runoff. It will enhance infiltration and, eventually, recharge the groundwater aquifer through rainfall and streamflow. The knowledge acquired through this research is helpful in understanding how the changes in land use influence strategies of water resources management and how the Ogun-Oshun River Basin Development Authority (OORBDA), stakeholders, and decision-makers can benefit from afforestation and regeneration for sustainable development in the ORB.

This research is an aspect of a Doctoral study sponsored by the German Ministry of Education and Research (BMBF), undertaken at the University of Abomey-Calavi, Republic of Benin as part of the West African Science Service Centre on Climate Change and Adapted Land Use (WASCAL) program on Climate Change and Water Resources.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.

Akintuyi
A. O.
,
Fasona
M. J.
,
Ayeni
A. O.
&
Soneye
A. S.
(
2021
)
Land use/land cover and climate change interaction in the derived savannah region of Nigeria
,
Environmental Monitoring and Assessment
,
193
(
12
),
1
24
.
Alabi
K.
,
Tobore
A.
,
Oyerinde
G.
&
Senjobi
B.
(
2021
)
Forest cover change in Onigambari reserve, Ibadan, Nigeria: application of vegetation index and Markov chain techniques
,
The Egyptian Journal of Remote Sensing and Space Science
,
24
(
3
),
983
990
.
Alademomi
A. S.
,
Okolie
C. J.
,
Daramola
O. E.
,
Akinnusi
S. A.
,
Adediran
E.
,
Olanrewaju
H. O.
,
Alabi
A. O.
,
Salami
T. J.
&
Odumosu
J.
(
2022
)
The interrelationship between LST, NDVI, NDBI, and land cover change in a section of Lagos metropolis, Nigeria
,
Applied Geomatics
,
14
(
2
),
299
314
.
Amankwah
A. A.
,
Quaye-Ballard
J. A.
,
Koomson
B.
,
Amankwah
R. K.
,
Awotwi
A.
,
Kankam
B. O.
,
Opuni-Frimpong
N. Y.
,
Baah
D. S.
&
Adu-Bredu
S.
(
2021
)
Deforestation in forest-savannah transition zone of Ghana: Boabeng-Fiema monkey sanctuary
,
Global Ecology and Conservation
,
25
,
e01440
.
Ashaolu
E. D.
,
Olorunfemi
J. F.
&
Ifabiyi
I. P.
(
2019
)
Assessing the spatio-temporal pattern of land use and land cover changes in Osun drainage basin, Nigeria
,
Journal of Environmental Geography
,
12
(
1–2
),
41
50
.
Asinwa, I. O., Olajuyigbe, S. O. & Adegeye, A. O. (2018) Tree species diversity, composition and structure in Ogun River Watershed, Southwestern Nigeria, Journal of Forestry Research and Management, 15 (1), 114–1134.
Awe
B. S.
,
Lasisi
M. O.
&
Omotayo
F. S.
(
2020
)
Analysis of rainfall characteristics and flood incidences in Ogun River Basin, Nigeria
,
International Journal of Research and Innovation in Applied Science (IJRIAS)
,
V
(
XII
), 119–126.
Awoniran
D. R.
,
Adewole
M. B.
,
Adegboyega
S. A.
&
Anifowose
A. Y. B.
(
2014
)
Assessment of environmental responses to land use/land cover dynamics in the Lower Ogun River Basin, Southwestern Nigeria
,
African Journal of Environmental Science and Technology
,
8
(
2
),
152
165
.
Ballut-Dajud
G. A.
,
Sandoval Herazo
L. C.
,
Fernández-Lambert
G.
,
Marín-Muñiz
J. L.
,
López Méndez
M. C.
&
Betanzo-Torres
E. A.
(
2022
)
Factors affecting wetland loss: a review
,
Land
,
11
(
3
),
434
.
Belgiu
M.
&
Dragut
L.
(
2016
)
Random forest in remote sensing: a review of applications and future directions
,
ISPRS Journal of Photogramming Remote Sensing
,
114
,
24
31
.
Boulesteix
A. L.
,
Janitza
S.
,
Kruppa
J.
&
König
I. R.
(
2012
)
Overview of random forest methodology and practical guidance with emphasis on computational biology and bioinformatics
,
Wiley Interdisciplinary Reviews: Data Mining and Knowledge Discovery
,
2
(
6
),
493
507
.
Casu
F.
,
Manunta
M.
,
Agram
P. S.
&
Crippen
R. E.
(
2017
)
Big remotely sensed data: tools, applications and experiences
,
Remote Sensing and Environment
,
202
,
1
2
.
Dey
N. N.
,
Al Rakib
A.
,
Al
K.
&
Raikwar
V.
(
2021
)
Geospatial modelling of changes in land use/land cover dynamics using multi-layer perception Markov chain model in Rajshahi City, Bangladesh
,
Environmental Challenge
,
4
,
100148
.
https://doi. org/10.1016/j.envc. 2021.100148
.
Dimobe
K.
,
Goetze
D.
,
Ouédraogo
A.
,
Forkuor
G.
,
Wala
K.
,
Porembski
S.
&
Thiombiano
A.
(
2017
)
Spatio-temporal dynamics in land use and habitat fragmentation within a protected area dedicated to tourism in a Sudanian Savanna of West Africa
,
Journal of Landscape Ecology
,
10
(
1
),
75
95
.
https:// doi. org/ 10. 1515/ jlecol- 2017- 0011
.
Dubertret
F.
,
Le Tourneau
F. M.
,
Villarreal
M. L.
&
Norman
L. M.
(
2022
)
Monitoring annual land use/land cover change in the Tucson metropolitan area with google earth engine (1986–2020)
,
Remote Sensing
,
14
(
9
),
2127
.
Eastman
J. R.
(
2020
)
TerrSet Geospatial Monitoring and Modeling System, Tutorial Version 2020v.19.0
.
Worcester
:
Clark University
.
Ewunetu
A.
,
Simane
B.
,
Teferi
E.
&
Zaitchik
B. F.
(
2021
)
Mapping and quantifying comprehensive land degradation status using spatial multicriteria evaluation technique in the headwaters area of upper blue Nile river
,
Sustainability (Switzerland)
,
13
(
4
),
1
28
.
https://doi.org/10.3390/su13042244
.
Falahatkar
S.
,
Soffianian
A. R.
,
Khajeddin
S. J.
,
Ziaee
H. R.
&
Nadoushan
M. A.
(
2011
)
Integration of remote sensing data and GIS for prediction of land cover map
,
International Journal of Geomatics and Geosciences
, 1 (4), 847–864.
Fashae
O. A.
,
Adagbasa
E. G.
,
Olusola
A. O.
&
Obateru
R. O.
(
2020
)
Land use/land cover change and land surface temperature of Ibadan and environs, Nigeria
,
Environmental Monitoring and Assessment
,
192
(
2
),
1
18
.
Gorelick
N.
,
Hancher
M.
,
Dixon
M.
,
Ilyushchenko
S.
,
Thau
D.
&
Moore
R.
(
2017
)
Google earth engine: planetary-Scale geospatial analysis for everyone
,
Remote Sensing Environment
,
2017
(
202
),
18
27
.
Guzha
A. C.
,
Rufino
M. C.
,
Okoth
S.
,
Jacobs
S.
&
Nóbrega
R. L. B.
(
2018
)
Impacts of land use and land cover change on surface runoff, discharge and low flows: evidence from East Africa
,
Journal of Hydrology: Regional Studies
,
15
,
49
67
.
Hackman
K. O.
,
Gong
P.
&
Wang
J.
(
2017
)
New land-cover maps of Ghana for 2015 using Landsat 8 and three popular classifiers for biodiversity assessment
,
International Journal of Remote Sensing
,
38
(
14
),
4008
4021
.
https:// doi. org/ 10. 1080/ 01431 161. 2017. 13126 19
.
Hackman
K. O.
,
Li
X.
,
Asenso-Gyambibi
D.
,
Asamoah
E. A.
&
Nelson
I. D.
(
2020
)
Analysis of geo-spatiotemporal data using machine learning algorithms and reliability enhancement for urbanization decision support
,
International Journal of Digital Earth
,
13
(
12
),
1717
1732
.
Hasan, S., Shi, W., Zhu, X., Abbas, S. & Khan, H. U. A. (2020) Future simulation of land use changes in rapidly urbanizing South China based on land change modeler and remote sensing data, Sustainability, 12 (11), 4350.
Hua
A. K.
(
2017
)
Land use land cover changes in detection of water quality: a study based on remote sensing and multivariate statistics
,
Journal of Environmental Public Health
, 2017 (1),
7515130
.
Hussien, K., Kebede, A., Mekuriaw, A., Beza, S. A & Erena, S. H. (2022) Modelling spatiotemporal trends of land use land cover dynamics in the Abbay River Basin, Ethiopia, Model Earth System Environment, 9, 347–376.
Idowu
O. O.
(
2013
)
Challenges of urbanization and urban growth in Nigeria
,
American Journal of Sustainable Cities and Society
,
2
(
1
),
79
94
.
Komolafe
A. A.
,
Awe
B. S.
,
Olorunfemi
I. E.
&
Oguntunde
P. G.
(
2020
)
Modelling flood-prone area and vulnerability using integration of multi-criteria analysis and HAND model in the Ogun River Basin, Nigeria
,
Hydrological Sciences Journal
,
65
(
10
),
1766
1783
.
Kulkarni
V. Y.
&
Sinha
P. K.
(
2013
)
Random forest classifiers: a survey and future research directions
,
International Journal of Advanced Computing
,
36
(
1
),
1144
1153
.
Lambin
E. F.
,
Geist
H. J.
&
Lepers
E.
(
2003
)
Dynamics of land-use and land-cover change in tropical regions
,
Annual Review of Environment and Resources
,
28
,
205
241
.
https://doi.org/10.1146/annurev.energy.28.050302.105459
.
Larbi
I.
,
Forkuor
G.
,
Hountondji
F.
,
Agyare
W.
&
Mama
D.
(
2019
)
Predictive land use change under business-as-usual and afforestation scenarios in the Vea catchment, West Africa
,
International Journal of Advanced Remote Sensing GIS
,
8
(
1
),
3011
3029
.
Liu
X.
,
Huang
Y.
,
Xu
X.
,
Li
X.
,
Li
X.
,
Ciais
P.
,
Lin
P.
,
Gong
K.
,
Ziegler
A. D.
,
Chen
A.
,
Gong
P.
,
Chen
J.
,
Hu
G.
,
Chen
Y.
,
Wang
S.
,
Wu
Q.
,
Huang
K.
,
Estes
L.
&
Zeng
Z.
(
2020b
)
High-spatiotemporal-resolution mapping of global urban change from 1985 to 2015
,
Nature Sustainability
,
3
(
7
),
564
570
.
https:// doi. org/ 10.1038/ s41893- 020- 0521-x (2020)
.
Midekisa
A.
,
Holl
F.
,
Savory
D. J.
,
Andrade-Pacheco
R.
,
Gething
P. W.
,
Bennett
A.
&
Sturrock
H. J. W.
(
2017
)
Mapping land cover change over continental Africa using Landsat and Google Earth Engine cloud computing
,
PLoS ONE
,
12
(
9
),
1
15
.
https://doi.org/10.1371/journal.pone.0184926
.
Mishra
V. N.
,
Rai
P. K.
&
Mohan
K.
(
2014
)
Prediction of land use changes based on land change modeler (LCM) using remote sensing: a case study of Muzaffarpur (Bihar), India
,
Journal of the Geographical Institute’ Jovan Cvijic’, SASA
,
64
(
1
),
111
127
.
Mmom
P. C.
&
Chukwu-Okeah
G. O.
(
2011
)
Factors and processes of coastal zone development in Nigeria: a review
,
Research Journal of Environmental and Earth Sciences
,
3
(
6
),
625
632
.
Na
X. D.
,
Zhang
S. Q.
,
Li
X. F.
,
Yu
H. A.
&
Liu
C. Y.
(
2010
)
Improved land cover mapping using random forests combined with landsat thematic mapper imagery and ancillary geographic data
,
Photogramming Engineering Remote Sensing
,
76
,
833
840
.
Nery
T.
,
Sadler
R.
,
Solis-Aulestia
M.
,
White
B.
,
Polyakov
M.
&
Chalak
M.
(
2016
). '
Comparing supervised algorithms in Land Use and Land Cover classification of a Landsat time-series
’,
‘2016 IEEE’ International Geoscience and Remote Sensing Symposium (IGARSS)
, pp.
5165
5168
.
IEEE
.
doi:10.1109/IGARSS.2016.7730346
.
Nugroho
P.
,
Marsono
D.
,
Sudira
P.
&
Suryatmojo
H.
(
2013
)
Impact of land-use changes on water balance
,
Procedia Environmental Sciences
,
17
,
256
262
.
Nut
N.
,
Mihara
M.
,
Jeong
J.
,
Ngo
B.
,
Sigua
G.
,
Prasad
P. V. V.
&
Reyes
M. R.
(
2021
)
Land Use and land cover changes and its impact on soil erosion in stung Sangkae catchment of Cambodia
,
Sustainability
,
13
(
16
),
9276
.
https://doi.org/10.3390/ su131 69276
.
Nzoiwu
C. P.
,
Agulue
E. I.
,
Mbah
S.
&
Igboanugo
C. P.
(
2017
)
Impact of land use/land cover change on surface temperature condition of Awka Town, Nigeria
,
Journal of Geographic Information System
,
9
(
06
),
763
.
Odunuga
S.
&
Oyebande
L.
(
2007
)
Change detection and hydrological implications in the lower ogun flood plain, SW Nigeria
,
Water and Energy Abstracts
,
17
(
4
),
12
12
.
Ogato
G. S.
,
Bantider
A.
&
Geneletti
D.
(
2021
)
Dynamics of land use and land cover changes in Huluka watershed of Oromia Regional State, Ethiopia
,
Environmental Systems Research
,
10
(
1
), 1–20.
https://doi.org/10.1186/s40068-021-00218-4
.
Ojo
S. A.
,
Olusina
J. O.
,
Ngene
B. U.
&
Olukanni
D. O.
(
2021
)
Modelling the relationship between climatic variables and land use/land cover classes in Yewa South Local Government Area of Ogun State, Nigeria
.
IOP Conference Series: Materials Science and Engineering
,
1036
(1),
012003
.
Oliveira
V. A. D.
,
Mello
C. R. D.
,
Viola
M. R.
&
Srinivasan
R.
(
2018
)
Land-use change impacts on the hydrology of the upper Grande River basin, Brazil
,
Cerne
,
24
,
334
343
.
doi:10.1590/01047760201824042573
.
Onanuga
M. Y.
,
Eludoyin
A. O.
&
Ofoezie
I. E.
(
2022
)
Urbanization and its effects on land and water resources in Ijebuland, southwestern Nigeria
,
Environment, Development and Sustainability
,
24
(
1
),
592
616
.
Oyerinde
G. T.
,
Hountondji
F. C.
,
Wisser
D.
,
Diekkrüger
B.
,
Lawin
A. E.
,
Odofin
A. J.
&
Afouda
A.
(
2015
)
Hydro-climatic changes in the Niger basin and consistency of local perceptions
,
Regional Environmental Change
,
15
(
8
),
1627
1637
.
Perez-Vega, A., Mas, J. F. & Ligmann-Zielinska, A. (2012) Comparing two approaches to land use/cover change modeling and their implications for the assessment of biodiversity loss in a deciduous tropical forest, Environmental Modeling Software, 29, 11–23.
Pontius
G. R.
(
2000
)
Quantification error versus location error in comparison of categorical maps
,
Photogrammetric Engineering Remote & Sensing
,
66
,
1011
1016
.
Prasomsup
W.
,
Piyatadsananon
P.
,
Aunphoklang
W.
&
Boonrang
A.
(
2020
)
Extraction technic for built-up area classification in Landsat 8 imagery
,
International Journal of Environmental Science and Development
,
11
(
1
),
15
20
.
Rimal, B., Zhang, L., Keshtkar, H., Wang, N. & Lin, Y. (2017) Monitoring and modeling of spatiotemporal urban expansion and land-use/land-cover change using integrated Markov chain cellular automata model, ISPRS International Journal of Geo-Information, 6 (9), 288.
Rindfuss
R. R.
,
Walsh
S. J.
,
Turner
B. L.
,
Fox
J.
&
Mishra
V.
(
2004
)
Developing a science of land change: challenges and methodological issues
,
Proceedings of the National Academy of Sciences
,
101
(
39
),
13976
13981
.
Roy
H. G.
,
Fox
D. M.
&
Emsellem
K.
(
2014
). '
Predicting Land Cover Change in a Mediterranean Catchment at Different Time Scales
’,
Proceedings of the International Conference on Computational Science and Its Applications
.
Guimaraes, Portugal
,
30 June 2014
, pp.
315
330
.
Roy
S.
,
Farzana
K.
,
Papia
M.
&
Hasan
M.
(
2015
)
Monitoring and prediction of land use/land cover change using the integration of Markov chain model and cellular automation in the Southeastern Tertiary Hilly Area of Bangladesh
,
International Journal of Sciences: Basic Applied Research
,
24
,
125
148
.
Setegn
S. G.
,
Srinivasan
R.
,
Dargahi
B.
&
Melesse
A. M.
(
2009
)
Spatial delineation of soil erosion vulnerability in the Lake Tana Basin, Ethiopia
,
Hydrological Processes: An International Journal
,
23
,
3738
3750
.
https://doi.org/10.1002/hyp.7476
.
Sloan
S.
,
Meyfroidt
P.
,
Rudel
T. K.
,
Bongers
F.
&
Chazdon
R.
(
2019
)
The forest transformation: planted tree cover and regional dynamics of tree gains and losses
,
Global Environmental Change
,
59
,
101988
.
Tadele
H.
,
Mekuriaw
A.
,
Selassie
Y. G.
&
Tsegaye
L.
(
2017
)
Land use/land cover factor values and accuracy assessment using a gis and remote sensing in the case of the Quashay watershed in Northwestern Ethiopia
,
Journal of Natural Resources and Development
,
7
,
38
44
.
https://doi.org/10.5027/jnrd.v7i0.05
.
Talukdar
S.
,
Singha
P.
,
Mahato
S.
,
Pal
S.
,
Liou
Y. A.
&
Rahman
A.
(
2020
)
Land-use land-cover classification by machine learning classifiers for satellite observations – A review
,
Remote Sensing
,
12
(
7
),
1135
.
Teluguntla
P.
,
Thenkabail
P. S.
,
Oliphant
A.
,
Xiong
J.
,
Gumma
M. K.
,
Congalton
R. G.
,
Yadav
K.
&
Huete
A.
(
2018
)
A 30-m landsat-derived cropland extent product of Australia and China using random forest machine learning algorithm on Google Earth Engine cloud computing platform
,
ISPRS Journal of Photogramming Remote Sensing
,
144
,
325
340
.
Thiam
S.
,
Salas
E. A. L.
,
Hounguè
N. R.
,
Almoradie
A. D. S.
,
Verleysdonk
S.
,
Adounkpe
J. G.
&
Komi
K.
(
2022
)
Modelling land use and land cover in the transboundary mono river catchment of Togo and Benin using Markov chain and stakeholder's perspectives
,
Sustainability
,
14
,
4160
.
https:// doi. org/ 10. 3390/ su140 74160
.
Tobore
A. O.
,
Senjobi
B.
&
Oyerinde
G.
(
2021
)
Spatio temporal analysis and simulation of land use and land cover change in Odeda peri- urban of Ogun State, Nigeria
,
Jordan Journal Earth Environmental Science (JJEES)
,
12
(
4
),
326
336
.
ISSN 1996–6681
.
Verburg
P. H.
,
Neumann
K.
&
Nol
L.
(
2011
)
Challenges in using land use and land cover data for global change studies
,
Global Change Biology
,
17
(
2
),
974
989
.
Xie
S.
,
Liu
L.
,
Zhang
X.
,
Yang
J.
,
Chen
X.
&
Gao
Y.
(
2019
)
Automatic land-cover mapping using landsat time-series data based on google earth engine
,
Remote Sensing
,
11
,
3023
.
Yalew
S. G.
,
Mul
M. L.
,
Griensven
A. V.
,
Teferi
E.
,
Priess
J.
,
Schweitzer
C.
&
Zaag
P.
(
2016
)
Land-use change modelling in the Upper Blue Nile Basin
,
Environments
,
3
(
3
),
21
.
Yangouliba
G. I.
,
Zoungrana
B. J. B.
,
Hackman
K. O.
,
Koch
H.
,
Liersch
S.
,
Sintondji
L. O.
,
Dipama
J-M.
,
Kwawuvi
D.
,
Ouedraogo
V.
,
Yabré
S.
,
Bonkoungou
B.
,
Sougué
M.
,
Gadiaga
A.
&
Koffi
B.
(
2022
)
Modelling past and future land use and land cover dynamics in the Nakambe River Basin, West Africa
,
Modeling Earth Systems and Environment
, 9 (2),
1651
1667
.
This is an Open Access article distributed under the terms of the Creative Commons Attribution Licence (CC BY 4.0), which permits copying, adaptation and redistribution, provided the original work is properly cited (http://creativecommons.org/licenses/by/4.0/).

Supplementary data