Introduction

Alongside population expansion and global warming, land use land cover change (LULCC) is a crucial component of environmental change on a worldwide scale. LULCC has recently gained international attention. Therefore, taking into account land use and land cover dynamics is a crucial component in achieving sustainable development with proper land use management and regional development1. All man-made and natural structures that encircle certain areas, such as flora and water bodies, are referred to as land cover. However, human activities that involve the soil can be used to determine land use. LULC change is one among the factors contributing to ecological changes at the local, regional, and global levels2. Over the past few decades, there has been a notable rise in the area of grassland, woodland, and forest that has been turned into farmland and grazing land3,4. On a typical year, 16 million hectares of the forests are destroyed worldwide and the trend is continually steady over time5.

In most of developing nations, land use land cover dynamics are rapidly rising, and the key factors to be cited are high economic growth and population increase coupled with globalization. As an illustration, drastic changes of natural landscapes took place in the eastern part of Africa due to human activities6, and Woodlands, grasslands, and vegetated spots have now been converted to agricultural land7. Unfavorable hydrological processes like runoff, infiltration, evapotranspiration, and rainfall patterns are also influenced by changes in land cover and land use brought on by urbanization, deforestation, and the growth of crop fields8,9,10,11,12. Soil moisture is also significantly influenced by the form and quantity of groundwater and surface discharge13.

Many scientific researches show that LULCC is deeply connected to the phenomenon of flooding and significantly affects flood phenomena dynamics14,15,16. Both natural and man-made factors may result in floods, and one of these factors is changes in land use and cover17. Dynamic changes in land cover made an area more susceptible to flooding.

Human invasions, whether direct or indirect, can significantly alter land use and cover18,19 and rapid population increase is a major contributing factor, particularly in nations like Ethiopia20. The LULC of Ethiopia has occurred in great distinction over the past decades due to overgrazing, urbanization, agriculture growth, and population expansion. The Ethiopian population is currently increasing rapidly with the population currently standing at 108 million people whose growth rate has been estimated to average 2.8% same as many other sub-Saharan African countries21.

The most important basin is the Awash River Basin, which is heavily used and provides water for industry, hydropower plants, households, and large and small-scale irrigation schemes. Despite this, Ethiopia’s Awash River Basin is environmentally endangered22,23,24,25. Because of its strategic location, accessibility to roadways, and the availability of land and water resources, the basin is susceptible to human-induced effects such as land degradation and high population density. Many Ethiopian scholars have studied changes in land use and land cover within the Awash River Basin, and their results lately revealed that there are significant changes in the area. Land degradation, water resource depletion, urbanization, and rapid population expansion are some of the primary issues facing the Awash River Basin24,26,27,28,29,30.

Land use and land cover change entails some of the challenges such as land degradation, urbanization, and water depletion as closely interconnected as knots, which strongly affect both the environment and population dynamics. Increased urban uses such as population growth often results in expansion of urban areas encroaching upon previously undeveloped land, which results in habitat loss, fragmentation, and land degradation. Although this expansion usually means that more resources are being extracted, especially water concentrated in urban centers, urban centers need significant supplies for drinking, sanitation, and industrial use. This can cause depletion of aquifers and surface water bodies, over extraction of water resource can also result in further degradation of the land and increase deterioration of the ecosystem. Furthermore, the change from natural landscapes to urban infrastructure affects local climate, reduces biodiversity, and affects the health of soil. A critical point of knowledge regarding these interactions is critical to the development of sustainable land use policies between urban development interest and environmental conservation as well as resource management. Therefore, it is crucial to comprehend the basin-level LULC dynamics and driving forces in order to establish and plan effective environmental management policies and strategies31.

The most popular technique for producing comprehensive land use classification maps is the use of remote sensing and geographic information system techniques, which are utilized to explore modifications in LULCC32,33,34,35,36. The rates, extents, and patterns of land-use dynamics can be ascertained by mapping land use changes using remote sensing techniques, which offer a quantitative description of land transformation. Numerous unique remote sensing categorization techniques have been developed and are frequently used to map changes in land cover and use37. One of the most popular classification techniques is supervised classification38,39,40. Maps of land use, land cover change, forest and other natural resource inventories, and other topics were produced in numerous studies using GIS and remote sensing technology22,26,41,42,43,44,45,46,47,48.

However, in the case of the Awash River Basin, the relationships between LULCC, population expansion, and flood dangers are rarely examined and discussed in detail. There is also a lack of understanding and practice regarding the introduction of nature-based solution (NBS) approaches. Hence, this study expected to examine the dynamics and relationship between land-use land cover change and population growth in relation to flood hazards in ARB through remote sensing and Geographic Information Systems. Moreover, suggest a sustainable nature-based solution to reduce the consequences of unexpected floods and rapid land use changes. Furthermore, this research helps to achieve the Sustainable Development Goals for 2030: Goal 15 encompasses investigating land use changes and conservation efforts to protect ecosystems and biodiversity.

Materials and methods

Study area description

Locations

The Awash River Basin is located between 7°53′ N–12° 0′ N latitude and 37°57′ E–43°25′ E longitude (Fig. 1). It is home to 10.5 million people and occupies an area of 112,000 km2, with an annual water surface area of about 4.9 BM349. The elevation above sea level varies from 210 to 4195 m. Parts of the Oromia, Somali, Amhara, and Afar regions are embraced by the 1250 km long Awash River29. Originating on a high plateau close to Ginchi town, west of Addis Ababa, Ethiopia’s capital, the river runs across the Rift Valley into the Afar Triad until spilling into the filthy Lake Abbe at the Ethiopian border with Djibouti50.

Based on physical and socioeconomic characteristics, the Awash River Basin is separated into three sections: the Middle Awash, which includes areas between Koka Dam and Awash Sebat station (areas between 1000 and 1500 m); the Lower Awash, which includes areas between 500 and 1000 m and the Eastern catchment; and the Upper Awash, which includes areas upstream from Koka Dam (areas above 1500 m)24,49. In upper and middle Awash, the rainfall pattern is bimodal, with the major rainy season occurring from July to September and an irregular wet season from March to May. The higher and lower regions of the Awash River Basin experience different climates, ranging from humid to desert. The Awash River Basin’s soil is classified as high clay soils that are good for cultivating cotton, clay loam and sandy loam for growing cereals and sugarcane, and sandy soil for range grass pasture49.

Fig. 1
figure 1

Study area map; the figure generated with GIS software version 10.8 by the Authors. The link to the GIS software used can be found at https://www.arcgis.com/index.html.

Full size image

Data collection and sources

For 1990, 2000, 2010, and 2020, four cloud-free Landsat TM and ETM images were obtained and thoroughly analyzed (Table 1). The patterns and trends of land use land cover changes in the study area were quantitatively measured and observed. All image bands 1–7 have a spatial resolution of 30 m. The images were obtained from the United States Geological Survey (USGS) through http://glovis.usgs.gov free of charge; they were geo-referenced to the Universal Transverse Mercator (UTM) map projection (Zone 37), WGS 84 datum. To observe clear changes in land use land cover, all satellite images from 1990 to 2020 were clipped or extracted from the same recent boundary shape file of the study area. During data collection for image classification and verification, we utilized GPS data to ensure accurate geo-location of agricultural areas and other land use types. Google Earth images provided high-resolution visual references for validating classifications, while field observations offered on-the-ground insights to enhance data quality and correct discrepancies. This comprehensive approach helped us to ensure the accuracy and reliability of our land-use classifications.

Table 1 Description of satellite data.
Full size table

Data analysis

Image classification

Determining the LULC classification algorithm is the most crucial stage in creating LULC maps. Six distinct LULC classes—urban area, agricultural land, forest cover, deep water bodies, bare land, and wetland—are defined for this study based on the standard classifications established by the US Geological Survey (Table 2). All satellite pictures are examined using spectral and spatial profiles prior to classification operations in order to verify the digital numbers (DNs) of various LULC classes. The training regions were selected using reference data and terrain maps. In ArcGIS 10.8, every satellite image is supervised and classified using a maximum likelihood approach. To compare supervised and unsupervised image classifications, according to Lu51, Choosing pixels that correspond to land cover classifications that the analyst is aware of are known as supervised classification. The classification of unsupervised images is more computer-automated. The use of unsupervised categorization may not produce satisfactory results because grass, crop, and shrub areas have similar spectral features and were previously identified as distinct classes. Consequently, supervised image classifications were employed in the data extraction procedure.

The reference data used in classifying the satellite images were the existing topographic maps, Google earth images and GPS coordinates of ground information. The classified images were used to obtain the areas of each of the LULC classes respectively in accordance with that of LULC scheme.

Table 2 Description of the land use land covers (LULC) types.
Full size table
Accuracy assessment

By evaluating commission and omission errors in a confusion or error matrix, the overall accuracy level is evaluated. In order to enhance classification methods and increase the quantity of overlooked misclassified cells, ArcGIS also uses classification and post-classification processes. It goes without saying that the anticipated errors in image classification may arise during the classification phase or earlier in the image processing step. Consequently, accuracy evaluation should be taken into consideration once classification concludes in order to raise the degree of reliability for image classification52. , Classification remains incomplete until accuracy assessment is done.

An error matrix or confusion matrix is one of the realistic widely used methods for evaluating classification accuracy53. Google Earth authentication and the collection of some GPS data were used to analyze the image classification accuracy. Sample points were collected from GPS coordinates and classified images corresponding to specific years. These points were converted into point features in ArcGIS environment and then visualized in Google earth by adjusting the time span to match the years of the classified images. Afterward, the points were compared against the respective images in Google earth for those specified years. This allows us to count and analyze the points that correspond accurately to the designated land use classes in the classified images over different time periods. The purpose of the confusion matrix is to examine and compare the ground truth points with the classification outcomes that were attained throughout the classification procedure54. A confusion matrix, which primarily uses various accuracy types, such as overall accuracy, user accuracy, producer accuracy, and the kappa coefficient (Eq. 1), was used to authenticate the categorized image in terms of percentage. The formulas that are briefly shown here were obtained from53,55,56,57.

$$\begin{aligned} {\text{Kappa}}\,{\text{ coefficient }}\left( {\text{T}} \right){\text{ }} = & \left( {{\text{TS}}*{\text{TCS}}} \right){\text{ }} - {\text{ }}\sum {\text{ }}\left( {{\text{column}}\,{\text{ total }}*{\text{ Row}}\,{\text{ total}}} \right)~ \times {\text{ 1}}00 \\ & {\text{TS}}^{{\text{2}}} - {\text{ }}\sum {\text{ }}\left( {{\text{column }}\,{\text{total }}*{\text{Row }}\,{\text{total}}} \right)~ \\ {\text{Overall }}\,{\text{Accuracy }}\left( \% \right){\text{ }} = & {\text{Total}}\,{\text{ number}}\,{\text{ of}}\,{\text{ correctly}}\,{\text{ classified }}\,{\text{Pixels}}\,{\text{ }}\left( {{\text{diagonal}}} \right)~ \times {\text{ 1}}00 \\ & {\text{Total }}\,{\text{number }}\,{\text{of }}\,{\text{reference}}\,{\text{ pixels }}\left( {{\text{sample}}} \right) \\ {\text{Users }}\,{\text{Accuracy }}\left( \% \right){\text{ }} = & {\text{Number}}\,{\text{ of}}\,{\text{ correctly}}\,{\text{ classified}}\,{\text{ Pixels}}\,{\text{ in}}\,{\text{ each}}\,{\text{ category}}~ \times {\text{ 1}}00 \\ & {\text{Total }}\,{\text{number }}\,{\text{of }}\,{\text{classified}}\,{\text{ pixels}}\,{\text{ in }}\,{\text{that }}\,{\text{category}}\,{\text{ }}\left( {{\text{the}}\,{\text{ row}}\,{\text{ total}}} \right) \\ {\text{Producers }}\,{\text{Accuracy }}\left( \% \right){\text{ }} = & {\text{Number}}\,{\text{ of}}\,{\text{ correctly}}\,{\text{ classified}}\,{\text{ Pixels }}\,{\text{in }}\,{\text{each }}\,{\text{category}} \times {\text{ 1}}00 \\ & {\text{Total }}\,{\text{number}}\,{\text{ of }}\,{\text{reference}}\,{\text{ pixels }}\,{\text{in }}\,{\text{that}}\,{\text{ category }}\,\left( {{\text{the }}\,{\text{column}}\,{\text{ total}}} \right) \\ \end{aligned}$$
(1)

where: TS = Total sample and TCS = Total Correctly classified Sample

Change detection

Various methods have been employed to identify changes in land cover and land use using remote sensing imagery. It ranges from detailed quantitative techniques to visual comparisons58.This paper incorporates post classification accuracy assessment, which is acknowledged as the most widely used and popular system59. From 1990 to 2020, changes in the research area were identified and examined. Determine which land use types have changed, to what extent, and in what percentages as compared to the state of the specified land use type in the prior year in order to assess the degree and scope of the changes. This kind of change study is crucial for figuring out how much change occurs. To determine the LULC %, subtract the initial LULC area from the final LULC area, divide the result by the original LULC area, and then multiply the result by 100 (Eq. 2)60.

$${\text{Change}}\,{\text{percentage}} = \frac{{\left( {{\text{Final}}\,{\text{~LULC}}\,{\text{~area~}} - {\text{~Initial}}\,{\text{~LULC}}\,{\text{~area}}} \right)}}{{{\text{~Initial~}}\,{\text{LULC}}\,{\text{~Area}}}}*100$$
(2)

The results of this matrix, which show a decline or rise in areal coverage in the study area, can be interpreted as negative as well as positive percentage values.

Methodological limitations

We had acquired vigorous data for LULCC analysis from remote sensing, but various limitations were recognized. The precision of remote sensing data was strongly impacted by atmospheric variables, sensor measurements, and varying terrain features. Moreover, during classification because of intersection in spectral signatures amid different terrain type possible biases may happen, consequently distort the classification. Hence, integration of high-resolution imagery and verification of ground truth comprising surveys in the field can significantly enhance the quality and data accuracy of future studies.

Population change analysis

Analyzing the study area’s population patterns was a crucial component for determining the primary sources of flood threats in the region with regard to LULCC assessment.

The Central Statistics Agency (CSA) and the Awash River Basin Authority provided the population census and projection statistics that served as the basis for the population data (Table 3). The population data for the years 1984, 1994, and 2007 were obtained from the Ethiopian Population and Housing Census61 and the literature62. The populations of the Awash River Basin in 1990, 2000, 2010 and 2020 were estimated via the population trend line from 1990 to 2020. In order to evaluate the effects of rising population on the dynamics of urbanization and land use change, the population also correlates with the pace of urban growth. The years that were chosen depended on the demographic data’s accessibility and ability to accurately depict changes in the population during the previous 30 to 40 years. A population’s change over a given time period is referred to as its population growth rate. According to the UN (2014)63, Ethiopia’s population is expected to rise by 2.3% annually, whereas urban regions’ average is 4.2%. The Ethiopian Central Statistical Agency’s equation (Eq. 3) is employed to generate projections61.

$${\text{P}}_{{\text{t}}} ~ = ~Po~e^{{r*t}}$$
(3)

where:

Pt : population projected at a given time,

Po: population size of a base year,

e: natural logarithm base,

r: average annual population growth rate, and.

t the time interval between the base year and the projected year.

Equation (3), obtained from the Ethiopian Central Statistics Agency, is used to get the estimated population in the Awash River basin, as seen in (Table 3).

Table 3 Awash river basin population estimates (in millions).
Full size table

The relationship between land-use/land-cover dynamics and population growth was quantitatively examined using the Pearson correlation coefficient (r). Finding the linear relationship between these two variables is the main goal of this analysis. A perfect positive correlation is represented by a coefficient (r) value of + 1, a perfect negative correlation by a value of -1, and no association by a value of zero64.

Thus, this approach enables us to understand the connections between shifts in land use patterns, including urbanization, agriculture, and deforestation, and shifts in population density and growth rates. The Pearson correlation can be used to find important connections and insights that help legislators, environmental managers, and urban planners better understand the relationship between ecological change and human activity. Equation (4) displays the Pearson correlation formula, which is frequently used.

$$r = \frac{{n(\sum xy) - (\sum x)(\sum y)}}{{\sqrt {\left[ {n\sum x^{2} - (\sum x)^{2} } \right]\left[ {n\sum y^{2} - (\sum y)^{2} } \right]} }}$$
(4)

where

r = Pearson coefficient.

n = number of pairs of the variables.

∑xy = sum of products of the paired variables.

∑x = sum of the x variables.

∑y = sum of the y variables.

∑x2 = sum of the squared x variables.

∑y2 = sum of the squared y variables.

The land consumption rate (LCR)

The LCR is a measure of a city’s degree of compactness that captures the city’s progressive spatial expansion. While a low value of LCR indicates the existence of open spaces, a high number suggests a high degree of crowdedness65. Researchers can comprehend the full effects of urban sprawl on cities’ socioeconomic and environmental conditions by combining LCR analysis with land-use dynamics. The approach assists experts in urban planning and policy in identifying trends that govern sustainable development, quality of life, and the management of natural resources.

Because metropolitan regions have more impermeable surfaces as a result of infrastructure development and urbanization, rainwater cannot rapidly sink into the ground after rainfall. Flooding follows from this. Therefore, assessing and understanding the rate of land consumption requires an understanding of land use-land cover change and flooding brought on by urbanization and growing population66,67,68,69. The land consumption rate, which was used as an indicator to measure how rapidly an urban area progressed geographically in relation to its population, was computed using Eq. (5).

$${\text{Land~}}\,{\text{consumption~}}\,{\text{rate}} = \frac{{{\text{~~Built}} - {\text{up}}\,{\text{~area~~~~~~~~~~~}}}}{{{\text{Population}}}}$$
(5)

Results and discussion

Accuracy valuation

The kappa coefficients for the same year image classifications were 0.92, 0.97, 0.99, and 0.95, respectively, while the overall accuracies for the 1990, 2000, 2010 and 2020 classifications were 94, 97.6, 99, and 96, respectively (Table 4). The accuracy of both the producer and the user in classifying the entire image met scientific requirements. According to57, if the kappa coefficient (K) value falls between 0.81 and 1.00, it’s nearly perfect, and the minimum total accuracy result should be 85%. As a result, this paper’s accuracy assessment level fulfills the necessary image classification threshold.

Land use land cover change rate

As is evident from Table 5; Fig. 2, in 1990, bare land made up the largest area coverage, accounting for 6,225,771 ha (53%) of the land use class. This was followed by agricultural land, which made up 2,768,467 Ha (24%). During the 1990–2020 study periods, agricultural land increased and eventually made up 23% of the land use class in 2020, while bare land slightly decreased and made up 26.5% of the land use class’s overall area coverage in 2020. Between 1990 and 2020, there was a notable and steady growth in the coverage of urban areas; for instance, in 1990, urban areas made up 22% of the land use class; in 2000, 2010, and 2020, that percentage rose to 26.2%, 30%, and 49%, respectively. The amounts of forest cover growing minimal between 1990 and 2020. Forest cover made about 0.7% of the study area in 1990 and rose dramatically to 1.5% of the river basin in 2000.

The amount of forest cover rapidly decreased from 0.5% in 2010 to 0.8% in 2020. Wetland and water bodies had the lowest area coverage, making up 0.5% and 0.12% of the total area coverage in 2020, respectively.

Regretfully, this kind of urbanization of barren and agricultural land can also result in a sufficient increase in impervious surfaces to raise the risk of flooding and runoff elsewhere. The hydrology of the basin may be affected by the changes, putting society at risk for more frequent and severe floods that would weaken community resilience and sustainability.

By improving water quality through natural filtration, expanding wetlands can boost biodiversity and serve as a buffer against the effects of flooding. However, there are significant negative effects of such metropolitan growth, including the potential for habitat fragmentation and pressure on local resources. The goal of this study is to give readers a more nuanced understanding of the interactions that occur within the ecosystem of the basin and the extent to which these interactions affect the communities that surround it.

Table 4 Accuracy assessment results for 1990, 2000, 2010 and 2020.
Full size table
Table 5 Total area coverage of land use classes.
Full size table

With reference to the change rate dynamics briefly described in (Table 6), there was a notable rise in agricultural land by 69% from 1990 to 2000. In contrast, bare land drastically declined by a 49% rate of change over the same year span. From 2000 to 2010 and 2010–2020, agricultural land significantly declined by 31% and 15%, respectively. In general, from 1990 to 2020, agricultural land declined by 1.3%; in contrast, bare land declined significantly, with a 50% rate of change within the same year range (1990–2020). This can be highly linked with an increasing rate of urbanization and population growth in the Awash River basin. Unfortunately, such conversion of agricultural land and bare land to urban areas can also increase impervious surfaces enough to enhance the potential for runoff and increased flood risk elsewhere. The changes could have an impact on the hydrology of the basin and place us at risk for more frequent and more severe flood events, which in turn would erode community resilience and sustainability.

Based on the findings of this paper, the primary reason for the decline in agricultural land is the conversion of agricultural areas into urban development, driven by significant population growth that has resulted in rapid urbanization. Additionally, environmental stresses such as climate change, drought, and soil degradation have negatively impacted agricultural productivity. As a result, there has been a decrease in the amount of arable land within the basin. Furthermore, fluctuations in commodity prices, limited access to markets, and high fertilizer costs also influence the extent of agricultural land that remains in cultivation.

According to Foeken (2006)70, food demand and food insecurity in cities are likely to rise in tandem with urbanization and population growth. Hence, to secure the existence of an urban community, urban irrigation practices, and intensive agricultural production take place as immediate remedies. In 2018, more than 250,000 ha of agricultural land were cultivated with extensive irrigation projects, which potentially increased during the coming decades49. As a result, in the future, we can expect a significant increase in agricultural land and a notable decline in bare land within densely populated and heavily utilized river basins in the country. This shift is driven by the high demand for food and the rising levels of food insecurity caused by population growth in urban areas.

The forest cover percentage rate significantly increased from 1990 to 2000 (112%) and declined by 66.6%. This reduction can be attributed to the unprecedented rate of deforestation, which has become a global issue. According to the FAO, Africa lost most of its forest between 1980 and 2000. Furthermore71, reported that rapid urban expansion was accompanied by a severe decline in green vegetation cover.

In addition, a slight increase of 60.5% from 2010 to 2020 occurred because of the expansion of afforestation and reforestation programs designed by the government. In general, from 1990 to 2020, the forest coverage change rate registered a positive value of 13.7%. The Green Legacy Initiative (GLI) of Ethiopia seeks to advance sustainable land management and reforestation. Implemented in the Awash River basin in Ethiopia, the GLI attempts to address environmental issues such as climate change, land degradation, and deforestation. Although the initiative’s overarching objectives are evident, its concrete effects on LULC changes and land surface temperature (LST) regulation are less clear. Moreover, in line with consideration of NDVI analysis, according to Gebremichael et al. (2025), the NDVI result in 2024 ranges from 1 to -0.2. The highest area coverage of the study area within this year shows significant greenness compared to other periods. In the Awash River basin, due to their high quantities of green biomass, crops, fruit trees, and forests are represented by high NDVI values, whereas urban areas and bare soil are represented by low NDVI values. More dense vegetation is indicated by higher NDVI values; this may have a strong relation with Green Legacy Initiatives (GLI), the reforestation program that started in 1999.

Another noteworthy shift pertains to the land use class of deep-water bodies, which experienced a 24% increase between 1990 and 2000, an 8.7% decline between 2000 and 2010, a significant decrease to 58% between 2010 and 2020, and a significant 43% reduction in deep water bodies of the Awash River basin between 1990 and 2020. Many studies have shown that urbanization results in changes in micro climates with large impacts on hydrological process, thermodynamic and solar radiation72. Moreover, some scholars have reported that the extraction of water from deep water bodies is highly practical in the Awash River Basin. According to Elleni (2009)73, small- and large-scale agricultural activity and the extensive activities of different industry developments are practiced in the Awash River Basin.

Moreover, towns such as Awash and Adama received their water supplies from the Awash River74. In addition, intensive water extraction from Lake Beseka (Metehara) and Haramaya for irrigation and the nearby city water supply has been highly demonstrated75. Scholar stressed that, in the Afar region, lakes and swamps have been fragmented in to small size spots. Moreover, since the 1930s, the extent of salty Lake Abbe has decreased by 67%.

Conversely, the wetland class has shown a significant increase of 191% during the previous thirty years (1990–2020). The wetland in the research area are mostly filled with floods and are susceptible to a possible increase in wetland coverage because of the frequent flash floods that occur in the basin riverbank area. In contrast41, reported that intensive utilization of wetland for irrigation practices would affect the sustainability of the basin. Furthermore75, argued that patches of wetland are converted into agricultural lands and that large water banks are renewed for hydroelectric power production or irrigation projects.

Another land use class that demonstrates notable and consistent changes is the urban area. From 1990 to 2000, the urban area change rate increased by 19%, and within the following two decades, significantly increased by 121% change rate. In the last three decades, rapid population growth has led to the extension of the development of urban areas and settlements in the study area. Many scholars agree with the link between population growth and urban expansion. In the basin of the Awash River from a socioeconomic standpoint, as the population and urban area grew, so did settlement areas and infrastructural development76. Roba et al. (2025) report that settlement areas increased gradually, adding 30.1 km² between 1993 and 2003 and then increasing by 3.1 km² after that, for a total gain of 33.1 km2. With a modest 6.0 km2 rise in the first ten years and a more substantial 35.4 km2 expansion in the following ten, sugarcane plantations gradually grew, reaching a total increase of 41.4 km². In general, the upper Awash River basin saw modest growth in settlement areas and sugarcane plantations26,76,77,78.

In other hand, regarding both land use class; wetland and urban area shows a substantial increment within the basin, expanding wetlands can increase biodiversity, improve water quality by natural filtration, and act as a buffer against flood influence. On the other hand, however, such expansions of urbanization have very huge consequences that can cause habitat fragmentation and put strain on nearby sources.

Ethiopia’s annual population growth rate is 2.3%; it is a higher rate than the developing countries, whose average rate is (2.2%), and this set Ethiopia as the second most populous country in Africa63. Eventually, rapid population growth accelerated the rate of urban expansion. Consequently, urbanization has a major influence on the abiotic environment79 and the whole living components of urban area80. Hence, this contributes to land use and land cover class changes, changes in the microclimate, surface temperature and rate of rainfall in the study area. Furthermore, owing to the high rate of population growth, the need for the consumption of natural resources has increased and has become a key challenge for the sustainable management of agriculture and forestland. The intensive consumption of forest resources and small patches of watersheds resulted in soil erosion in the lower stream of the Awash River Basin25.

In order to magnify the significant interrelation of LULCC, population growth, and flood hazard, the flood hazard map of the Awash River Basin was illustrated (Fig. 3). ArcGIS 10.8 was used to create, reclassify, and weight flood-causing elements, including slope, rainfall, elevation, LULC, drainage density, proximity to a river, and soil type, in order to create a flood hazard map. Lastly, utilizing AHP techniques, a flood hazard area map of the Awash River Basin (ARB) was created after weighted overlay. Additionally, the results of the classification of the flood hazard map produced in this work are quantitatively validated using a widely used validation technique: the area under the curve (AUC) of the Receiver Operating Characteristic (ROC) methods. The AUC score of 0.92 indicated an outstanding validation outcome. Population growth and general land use changes, including increased agricultural output, increased overflow, and created a high flood hazard zone in the downstream regions of Awash River basins. Recurrent floods have therefore increased in frequency each year in the Awash River Basin, causing large financial losses as well as associated societal effects. The primary cause of this is the climate and other associated factors. Additionally77, Gebremichael et al. (2025), emphasize that the majority of the basin, or 5,855,130.12 ha (51.4%), is at high risk of flooding, while the second-highest area coverage, or 4,625,378.51 ha (40.6%), of the entire Awash River Basin area coverage, is at moderate risk of flooding.

Table 6 Land use class change detection percentage rate.
Full size table

Relationship between population growth and LULCC using Pearson correlation

The Pearson correlation calculation indicates that the r values (Table 7 and Appendix 1) for bare land, deep water bodies, agricultural land, and forest coverage (− 0.75, − 0.49, − 0.28, & − 0.25, respectively) have a negative interaction with population growth. Thus, while the population of the Awash River basin is increasing at an astonishing rate, the aforementioned land use groups are seeing a sharp decline. Specifically, we can state with simplicity that when population growth increased significantly, the study area’s bare land transitioned significantly into different land use groups. In contrast, urban areas and wetlands are positively correlated with population growth (r values of 0.94 and 0.89, respectively). We can demonstrate how urbanization and population growth are substantially correlated. As a result, rural and semi-urban regions grew into urban areas, which in turn caused considerable changes in land use and cover. The strong positive correlation of urban and population growth shows that, the area of the urban tends to increase as the population increases that could add more stress on the already existing infrastructure and resources. Unsurprisingly, this could create problems like traffic congestion, higher demand for public services and heightened pressures on housing markets. They can help urban planners set the baselines for analyzing the sustainable strategies of growth determined by the scenario of projected population dynamics. For instance, it could entail focusing on the development of green spaces, the best land use strategy to avoid urban sprawl, and guaranteeing people’s access to essential services that will respond to their increasing requirements. Results and influence of the correlation on decision making processes concerning the development and management of resources in the region is highly significant.

According to81,82 the increase in the number of people and built-up patterns has caused alterations in land use–land cover and regularly results in flooding. This finding is also supported by62, who revealed that urbanization significantly correlates positively with population growth in the Awash River Basin. According to83 climate change, population growth and urbanization are among the driving forces behind floods.

On the other hand, the expansion of irrigation activities along the river line of the Awash River basin has also strongly increased, and the development of irrigation activities in the area led to the creation of various small patches of wetlands. The small wetlands in the study area were created due to natural processes, such as sedimentation, erosion and, changes in water flow patterns.

Moreover, frequent flash floods on the river side line are extremely common and contribute to the creation of patches of wetlands. Hence, these factors caused the wetland coverage in the Awash River basin to account for significantly greater coverage in association with population growth.

Table 7 Pearson correlation between different classifications of land use.
Full size table

Land consumption rate (LCR)

In 1990, 2000, 2010, and 2020, the Awash River Basin’s estimated land consumption rates (LCR) were 0.29, 0.28, 0.26, and 0.34, respectively (Table 8). Given the strong correlation between urbanization and the LCR, the rate has been gradually declining over the last thirty years, suggesting that the Awash River Basin has a low development density. However, the LCR decline rate in the studied area from 1990 to 2010 indicates that urban area growth is not keeping up with population increase. In the research area, there were still modest rates of population increase and urban area increment, nonetheless. According to65 the primary drivers of land use change and the increase of built-up land are population growth and developmental activity.

On the other hand, the LCR has risen sharply since 2020, which has somewhat accelerated the rise in progressive urbanization that is, the growth of urban areas has outpaced the rate of population expansion. According to84 there have been instances of extremely high LCR in certain cities, which can occasionally exceed double the population growth rate (PGR). Resulting in a high LCRPGR index, this demonstrates notable inefficiencies in urban land utilization and poor land resources management. Hence, the maximum urban area expansion was registered during 2020, and the trend also indicated that the rate will increase in the near future. This is considered urban extension or urban sprawl85. This finding is also supported by the finding from the Pearson correlation analysis of this study, which illustrates a positive association between land consumption rate and urban expansion.

The Awash River basin saw the fastest rate of LCR from 2010 to 2020, with significant LCR recorded throughout that time despite a very slow pace of population expansion. Inefficient urban land use and inadequate land resource management in the basin’s urban and peri-urban zones may be the primary cause of this. According to84 between 2007 and 2014, a rapid pace of LCR occurred in certain cities, including Addis Ababa. Others saw significant growth following 2014.

While population growth rate (PGR) is extremely low, a notable increase in LCR could be a hint of a rush by state or non-state entities to seize Peri-urban territory. In a similar vein, the cities’ LCRPGR index, this gauges urban land use efficiency. From 2014 to 2019, the majority of regional cities had high LCRPGR. Conversely, between 2007 and 2014, Addis Ababa’s outlying cities had a higher LCRPGR index.

Table 8 Land consumption rate for Awash river basin.
Full size table

This could suggest that Addis Ababa and the surrounding area experienced rapid urbanization and extensive rural land conversion before the regional cities were affected. Moreover, the majority of Ethiopian cities have low urban land use efficiency. Addis Ababa’s urban land efficiency is poorer than that of its regional peers, with an LCRPGR rating of 1.2: Nairobi (0.9), Dares Salaam (0.2), Khartoum (0.9), Kampala (0.4) and Kigali (0.6)86.

Fig. 2
figure 2

LULCC of Awash River Basin; the figure generated with GIS software version 10.8 by the Authors. The link to the GIS software used can be found at https://www.arcgis.com/index.html.

Full size image

Sustainable development goal: goal 15

Our research identifies critical factors influencing biodiversity and ecosystem health in the context of changing land use patterns. By analyzing the interactions between human activities and ecological systems, our findings reveal areas where conservation efforts can be most effectively focused to protect and restore biodiversity.

This alignment with Sustainable Development Goal: Goal 15 is paramount, as it aims to protect, restore, and promote sustainable use of terrestrial ecosystems and manage forests sustainably in Awash River Basin.

Moreover, our study suggests the need for integrating ecological considerations into land use planning processes to mitigate detrimental impacts on habitats and species. By advocating for sustainable practices such as agro-ecology, habitat preservation, and the restoration of ecological corridors, we aim to provide actionable recommendations that contribute toward achieving Sustainable Development Goal: Goal 15.

Fig. 3
figure 3

Flood hazard map of Awash River basin: the figure generated with GIS software version 10.8 by the Authors. The link to the GIS software used can be found at https://www.arcgis.com/index.html.

Full size image

Conclusion and recommendations

The main objective of this research paper is to analyze the dynamics of land-use land cover change and population growth with respect to flood hazards in the Awash River Basin using remote sensing and Geographic information system (GIS). Supervised classification for image from 1990, 2000, 2010 and 2020 was accurately conducted, and the results were effectively discussed. According to the findings from the land use land cover classifications, in the year 1990 in ARB, the area with highest coverage was bare land, followed by agricultural land, which was drastically transformed into different land use types. In contrast, urban area coverage significantly and consistently increased from 1990 to 2020, which was strongly linked with increasing rates of urbanization and population growth in the Awash River basin.

The findings of the Pearson correlation analysis indicate a favorable association between the expansion of the human population and both urban areas and wetlands. This connection might show the strong relationship between population growth and urbanization. Furthermore, the study demonstrates that while urban and population growth has expanded dramatically, the rate of land consumption in terms of density has remained negligible. In 2020, there was a noticeable increase in the study area’s population, urbanization, and land consumption rate. Since then, the growth of the population in conjunction with urban regions has the ability to impact land use land cover, which has led to flooding. Considering sustainable nature-based solutions to mitigate spontaneous rapid land use changes and flood impacts is invaluable. Some nature-based solutions that can be implemented in the region to improve the effectiveness of our recommendations are proposed. Wetland restoration is a very important aspect of improving natural habitat modification as it can help to both increase natural water retention and decrease flood risks and improve biodiversity. Additionally, reforestation can offer a huge pool to stabilize soil and regulate water. Enhanced watershed management and reduced surface runoff by increasing the forested areas of strategic locations mean that flood risks can be minimized. Promoting agro-forestry practices, such as triangular farming that combines agricultural productivity with ecological benefits through the establishment of resilient landscapes for the purpose of supporting local communities and environmental health would also be beneficial. The effect of incorporating these nature-based solutions does not only address immediate concerns had with land use changes and flooding but also greatly aligns with broader sustainability goals for the region and achieve long term resiliency in the region. Sustainable stakeholder actions can be proactively adopted by adopting the pragmatic, nature based ways of addressing social and economic needs while safeguarding ecological integrity.

The paper concludes that this study has significant implications for policy makers, urban and environmental planners. They highlight the importance of addressing the correlation between population growth, urbanization, and land-use land-cover changes to mitigate flood hazards. Reforestation initiatives, such as Ethiopia’s GLI, are successful in improving local microclimates, reversing land degradation, and restoring ecosystem services, all of which improve human health and agricultural output. This study also emphasizes the need for sustainable solutions that prioritize nature-based approaches to mitigate the impacts of severe land use land cover change and flood events. Overall, this research provides valuable insights into the complex relationships among land use land cover changes, population growth, and flood hazards in the Awash River Basin. Moreover, this research helps to succeed Sustainable Development Goal number 15 by 2030.