LAND USE CHANGE AND PLANNING COMPLIANCE IN LAGOS: ASSESSING ZONING RULES AND DRIVERS OF NON-COMPLIANCE

Volume-08 Issue 10, October2024 ISSN: 2456-9348 Impact Factor: 7.936 International Journal of Engineering Technology Research Management (IJETRM) https://ijetrm.com/ IJETRM (http://ijetrm.com/) [91] LAND USE CHANGE AND PLANNING COMPLIANCE IN LAGOS: ASSESSING ZONING RULES AND DRIVERS OF NON-COMPLIANCE Kayode A. Adeparusi Department of Earth, Atmospheric, & Geographic Information Sciences (EAGIS) Western Illinois University, USA. [email protected] ABSTRACT Rapid urbanization in Lagos has produced profound shifts in land use, often diverging from officially designated zoning rules. This study investigates the nature and extent of land use change between 2000 and 2025 and examines the socio-economic, institutional, and spatial drivers of zoning non-compliance. Using supervised classification of Landsat 8 and 9 and Sentinel-2 time series, post-classification change detection techniques, and overlays with official zoning maps and building permit records, we identify both compliant and non-compliant land developments. Spatial econometric models are employed to analyze the determinants of non-compliance across Lagos’ local government areas. Findings reveal that land use conversion from wetlands and peri-urban agricultural zones into residential and mixed-use developments is the dominant trend, with nearly 42% of observed developments occurring outside zoning allocations. Drivers of non-compliance include weak enforcement capacity, high population density, proximity to transport corridors, and speculative land markets. The study underscores the critical need for more adaptive urban planning frameworks, enhanced enforcement, and the integration of remote sensing and geospatial monitoring in policy implementation. Keywords: Land use change, zoning compliance, Lagos, supervised classification, spatial econometrics, urban planning 1. INTRODUCTION 1.1 Background and Rationale Urban land use planning is a cornerstone of sustainable city management. Zoning regulations, which specify the allocation of land to residential, commercial, industrial, and conservation uses, are designed to promote orderly development, mitigate conflicts between incompatible activities, and safeguard environmental resources. However, in many rapidly urbanizing African cities, zoning compliance is undermined by informal settlements, weak governance, and speculative pressures on land markets [1,2]. Lagos, Nigeria’s commercial hub, epitomizes the paradox of rapid urban growth amid fragile planning enforcement. With an estimated population exceeding 20 million residents, Lagos has expanded dramatically over the past three decades [3]. Driven by rural-urban migration, natural population increase, and economic concentration, the city’s built-up area has sprawled into wetlands, agricultural hinterlands, and ecologically sensitive zones [4]. Official zoning frameworks, however, continue to designate large portions of land for conservation, low-density residential use, or restricted industrial purposes. The disjuncture between planning prescriptions and actual land use outcomes raises urgent questions about the drivers and consequences of non-compliance. Recent advances in Earth observation technologies, particularly the availability of free high-resolution satellite imagery from the Landsat and Sentinel missions, have revolutionized the ability to monitor land use change at scale. Coupled with spatial econometric models, such data can illuminate not only what changes are occurring but also why. This is especially critical in Lagos, where informal settlements and unregulated conversions constitute a significant share of urban development [5,6]. 1.2 Urbanization and Zoning in Lagos Volume-09 Issue 10, October-2025 ISSN: 2456-9348 Impact Factor: 8.232 International Journal of Engineering Technology Research Management (IJETRM) https://ijetrm.com/ IJETRM (http://ijetrm.com/) [92] Zoning as a planning instrument in Lagos dates back to the colonial period, when planning ordinances established separate quarters for Europeans and Africans [7]. Post-independence, the Lagos State Urban and Regional Planning Board codified zoning regulations into master plans, notably the Lagos Metropolitan Master Plan (1980–2000) and subsequent development frameworks [8]. These plans envisioned a polycentric city structure, with residential, industrial, and commercial zones clearly delineated. In practice, however, compliance with zoning prescriptions has been inconsistent. High demand for affordable housing has led to the proliferation of informal settlements in areas zoned for agriculture or open space, while commercial developments frequently encroach on residential districts [9]. Weak enforcement mechanisms, inadequate resources within planning authorities, and political interference further exacerbate non-compliance [10]. At the same time, speculative land acquisition and rent-seeking behavior by both elites and local landholding families contribute to spatial outcomes that diverge sharply from official zoning [11]. The ecological implications are profound. Lagos’ wetlands, which provide flood regulation and biodiversity services, have been extensively reclaimed for housing estates, shopping complexes, and road infrastructure [12]. Such land use change increases flood vulnerability, reduces ecosystem resilience, and undermines long-term sustainability [13]. Understanding the extent and drivers of these patterns is essential to informing adaptive and enforceable planning strategies. 1.3 Research Problem and Objectives Despite the importance of zoning to urban governance, there is limited empirical research that systematically links observed land use change with planning compliance in Lagos. Previous studies have examined land cover dynamics using remote sensing [14,15] and have documented the proliferation of informal housing [16], but few have explicitly assessed zoning non-compliance relative to official maps. Similarly, while socio-economic drivers of urban growth have been analyzed [17], the integration of zoning overlays and econometric models of compliance remains underexplored. This study addresses these gaps by posing the central research question: How has land use in Lagos changed relative to zoning rules, and what drivers explain zoning non-compliance? To answer this question, the study sets out four objectives. One objective is to classify land use in Lagos for selected years using Landsat 8/9 and Sentinel-2 data. Another is to quantify land use changes through post-classification change detection. In addition, the research seeks to assess the extent of compliance with zoning allocations by overlaying classified maps with official zoning layers and building permit records. A further objective is to model the socioeconomic, institutional, and spatial drivers of zoning non-compliance using spatial econometric methods. 1.4 Significance of the Study This research offers both academic and policy contributions. Academically, it integrates remote sensing, GIS, and spatial econometric modeling into the study of zoning compliance, thereby advancing methodological approaches to urban land use analysis. Unlike descriptive studies of informal settlements, this study provides a systematic framework for linking actual land use with planned zoning allocations. From a policy perspective, findings can inform Lagos State planning authorities and policymakers in Nigeria more broadly. By identifying the drivers of non-compliance, the study provides an evidence base for designing targeted interventions, whether strengthening enforcement capacity, improving access to affordable housing, or revising outdated zoning frameworks. Moreover, the approach can be replicated in other rapidly urbanizing African cities facing similar challenges of compliance and enforcement. 2. METHODS 2.1 Study Area Lagos State, located in southwestern Nigeria, is the country’s economic hub and one of the fastest growing megacities in the world. Covering approximately 3,577 km², Lagos has an estimated population of over 20 million people and is characterized by rapid urbanization, high population density, and extensive informal settlements. The city’s coastal location and wetland ecosystems make it particularly vulnerable to environmental pressures such as flooding and land degradation [16]. 2.2 Data Sources Volume-09 Issue 10, October-2025 ISSN: 2456-9348 Impact Factor: 8.232 International Journal of Engineering Technology Research Management (IJETRM) https://ijetrm.com/ IJETRM (http://ijetrm.com/) [93] This study combined remote sensing datasets, official zoning layers, building permits, and socio-economic indicators. Landsat 8 and 9 OLI/TIRS multispectral data, with a spatial resolution of 30 m, were acquired at five-year intervals between 2013 and 2023. These datasets were selected for their long temporal coverage and consistent calibration, which enable historical comparisons [17]. To complement these, Sentinel-2 MSI imagery with 10 m resolution was obtained for the period 2016–2025, providing higher spatial detail that supports the validation of Landsat classifications [18]. Administrative datasets were obtained from the Lagos State Ministry of Physical Planning and Urban Development (MPP&UD). Official zoning maps delineating residential, commercial, industrial, agricultural, conservation, and mixed-use zones were digitized into a GIS database [19]. Building permit records for the years 2010–2024 were also collected to differentiate between authorized and unauthorized developments, though coverage was incomplete [20]. Socio-economic data included population density, poverty indices, and housing demand statistics, sourced from the Nigerian National Bureau of Statistics (NBS) and the Lagos State Bureau of Statistics [21]. Transport infrastructure data, including road networks and transit hubs, were acquired from OpenStreetMap and cross-validated against datasets provided by the Lagos Metropolitan Area Transport Authority (LAMATA) [22]. 2.3 Image Preprocessing All satellite imagery underwent preprocessing to ensure comparability across sensors and time periods. Atmospheric correction was applied using the LEDAPS algorithm for Landsat data and the Sen2Cor processor for Sentinel-2 data [23]. Geometric correction was performed by projecting all images to WGS84/UTM Zone 31N, maintaining a root mean square error (RMSE) of less than 0.5 pixels. Clouds and shadows were masked using the Fmask algorithm, ensuring that only cloud-free pixels were retained [24]. To achieve full coverage of Lagos, multiple satellite scenes were mosaicked for each study year, creating seamless datasets for classification. 2.4 Land Use Classification A six-class land use/land cover (LULC) scheme was employed, aligned with Lagos’ zoning categories. These included residential/urban built-up, commercial/industrial, agricultural land, wetlands/mangroves, open space/conservation, and water bodies. Training data were generated using stratified random sampling from multiple sources: highresolution Google Earth imagery between 2013 and 2023, ground-truth points collected during 2023 field surveys (n = 450), and digitized polygons of verified land uses from planning documents. Validation samples (n = 600) were withheld from training for independent accuracy assessment. Classification was carried out using the Random Forest (RF) algorithm, which has been shown to perform well in heterogeneous urban environments [25]. Model parameters were optimized through 10-fold cross-validation, setting the number of trees to 500 and adjusting maximum depth to minimize out-of-bag error. To harmonize results across sensors, Sentinel-2 imagery was resampled to 30 m and merged with Landsat classifications using a majority filter. Accuracy was then assessed using confusion matrices, reporting overall accuracy, producer’s accuracy, user’s accuracy, and the Kappa coefficient. Classifications with less than 85% overall accuracy were retrained using additional samples [26]. 2.5 Post-Classification Change Detection Land use change was quantified through post-classification comparison between time periods (2013–2018, 2018– 2023). Transition matrices were generated to identify dominant conversions. The annual rate of change was calculated as: R = At2 – At2 ×100 At1 × (t2 – t1) where At1 and At2 represent land use area at times t1 and t2. Spatial hot-spot analysis (Getis-Ord Gi*) was used to identify clusters of intense land use change. 2.6 Zoning Compliance Assessment The classified land use maps were overlaid with official zoning maps to assess compliance. A development was considered compliant when its observed land use class matched the designated zoning category, such as residential uses within residential zones. Non-compliance was defined as mismatches, for example, residential or commercial Volume-09 Issue 10, October-2025 ISSN: 2456-9348 Impact Factor: 8.232 International Journal of Engineering Technology Research Management (IJETRM) https://ijetrm.com/ IJETRM (http://ijetrm.com/) [94] development in areas zoned for agriculture or conservation. Building permit records were also integrated to differentiate between authorized and unauthorized developments. Compliance rates were calculated at both the state level and the Local Government Area (LGA) scale to highlight spatial variation [27]. 2.7 Spatial Econometric Modeling To investigate the drivers of zoning non-compliance, spatial econometric models were estimated. The dependent variable was the proportion of non-compliant land within each LGA, calculated as the ratio of non-compliant area to total land area. Explanatory variables included population density (expected positive), median household income (expected negative), distance to major roads (expected negative), proximity to the central business district (expected negative), presence of wetlands (expected positive), permit approval rate (expected negative), and a political influence score representing elite-dominated LGAs (expected positive) [28]. Three model specifications were tested. An Ordinary Least Squares (OLS) regression provided the baseline. A Spatial Lag Model (SLM) accounted for spillover effects across LGAs, while a Spatial Error Model (SEM) captured unobserved spatially correlated error terms. Spatial weight matrices were constructed using the queen contiguity method at the LGA level. Model selection was guided by Akaike Information Criterion (AIC) values and loglikelihood scores [29]. 2.8 Model Diagnostics Diagnostic tests were conducted to validate the models. Variance inflation factors (VIF) were used to assess multicollinearity, applying a threshold of less than 5. Heteroskedasticity was tested using the Breusch–Pagan test, while residual spatial autocorrelation was checked using Moran’s I statistic. Robustness was further tested by reestimating models with alternative spatial weight matrices, specifically using k-nearest neighbors with k set to 4 [30]. 2.9 Ethical Considerations This study relied primarily on publicly available remote sensing and administrative data. Ground-truth data collection involved field visits but did not include personal identifiers or household surveys. 3. RESULTS 3.1 Classification Accuracy The supervised classifications of Landsat and Sentinel imagery achieved strong performance across all time periods. Table 1 summarizes the accuracy metrics for 2013, 2018, and 2023 classifications. Table 1. Classification accuracy metrics (2013–2023). Year Overall Accuracy (%) Kappa Coefficient Producer’s Accuracy Range (%) User’s Accuracy Range (%) 2013 87.3 0.82 80.1–91.6 79.4–92.8 2018 88.9 0.84 82.5–93.2 81.7–94.0 2023 90.4 0.86 83.4–94.7 82.1–95.6 Wetland and agricultural classes showed slightly lower accuracies due to spectral confusion during the rainy season, but accuracies were consistently above the 80% threshold recommended for reliable land cover mapping. Built-up and water classes had the highest accuracies across all years, supported by strong spectral separability. 3.2 Land Use Change Trends 3.2.1 Overall Land Use Dynamics Between 2013 and 2023, Lagos experienced substantial land use transformation (Figure 1). Built-up areas expanded from 760 km² in 2013 to 1,085 km² in 2023, representing a 42.8% increase in a decade. Agricultural land declined sharply from 960 km² to 655 km², while wetlands shrank from 285 km² to 173 km². Conservation areas were encroached upon, decreasing by 19.4%. Water bodies remained relatively stable due to natural constraints. Volume-09 Issue 10, October-2025 ISSN: 2456-9348 Impact Factor: 8.232 International Journal of Engineering Technology Research Management (IJETRM) https://ijetrm.com/ IJETRM (http://ijetrm.com/) [95] Figure 1. Land use maps of Lagos for 2013, 2018, and 2023 derived from supervised classification. 3.2.2 Dominant Conversions Transition matrices (Table 2) reveal the major land use conversions. Table 2. Major land use transitions, 2013–2023 (km²). From → To Residential/Commercial Agricultural Wetlands Conservation Agricultural 242 – 16 47 Wetlands 65 7 – 12 Conservation 88 23 9 – The largest single conversion was agricultural to residential/commercial, accounting for 242 km² (36.9% of all change). Wetlands were disproportionately converted to residential estates and industrial complexes, reflecting widespread reclamation. Conservation areas were also converted to both residential and agricultural uses. 3.2.3 Spatial Hotspots of Change Hotspot analysis (Figure 2) identified three clusters of intense change: 1. Lekki Peninsula – rapid residential and commercial expansion into wetlands and conservation zones. 2. Alimosho and Agege LGAs – agricultural land converted into dense housing estates. 3. Badagry corridor – peri-urban sprawl driven by highway construction. Figure 2. Hotspot map of land use change intensity, 2013–2023 (Getis-Ord Gi). Volume-09 Issue 10, October-2025 ISSN: 2456-9348 Impact Factor: 8.232 International Journal of Engineering Technology Research Management (IJETRM) https://ijetrm.com/ IJETRM (http://ijetrm.com/) [96] 3.3 Zoning Compliance Assessment 3.3.1 Extent of Compliance Overlay analysis revealed that 41.7% of new developments between 2013 and 2023 occurred in zones inconsistent with official zoning allocations (Table 3). Table 3. Zoning compliance rates, 2013–2023. Development Type Compliant (%) Non-Compliant (%) Residential 62.4 37.6 Commercial/Industrial 54.9 45.1 Mixed-use 57.8 42.2 Commercial and industrial developments showed the highest levels of non-compliance, often located in residential or conservation zones. Residential encroachment into agricultural and conservation zones was also widespread. 3.3.2 Spatial Variation in Non-Compliance Non-compliance varied markedly across LGAs (Figure 3). • High non-compliance (>50%): Eti-Osa, Ibeju-Lekki, Amuwo-Odofin, Badagry. • Moderate non-compliance (30–50%): Alimosho, Ikorodu, Surulere. • Low non-compliance (<30%): Ikeja, Lagos Island, Apapa. Peripheral LGAs exhibited higher non-compliance due to weak enforcement and rapid peri-urban growth. Figure 3. Zoning compliance map by LGA, 2023. 3.3.3 Building Permit Analysis Cross-checking with building permit records revealed that only 58% of compliant developments had official permits, while 74% of non-compliant developments were unpermitted. This highlights the limited reach of formal planning approval processes, especially in peripheral LGAs. 3.4 Econometric Model Results 3.4.1 Ordinary Least Squares (OLS) The baseline OLS regression (Table 4) indicated significant positive associations between population density, proximity to roads, wetland presence, and non-compliance rates. However, spatial autocorrelation tests revealed significant clustering of residuals (Moran’s I = 0.219, p < 0.01), suggesting that OLS was misspecified. Volume-09 Issue 10, October-2025 ISSN: 2456-9348 Impact Factor: 8.232 International Journal of Engineering Technology Research Management (IJETRM) https://ijetrm.com/ IJETRM (http://ijetrm.com/) [97] Table 4. OLS regression results for drivers of zoning non-compliance. Variable Coefficient Std. Error t-stat p-value Population density 0.014 0.004 3.52 0.001 Median income -0.008 0.003 -2.69 0.011 Distance to major roads -0.021 0.006 -3.36 0.002 Distance to CBD -0.007 0.005 -1.44 0.163 Wetland presence (dummy) 0.118 0.042 2.81 0.009 Permit approval rate -0.036 0.012 -3.00 0.006 Political influence score 0.051 0.019 2.68 0.012 Adjusted R² = 0.41. 3.4.2 Spatial Lag Model (SLM) The SLM (Table 5) improved model fit (AIC reduction of 22 points) and showed a significant positive spatial lag coefficient (ρ = 0.26, p < 0.01), confirming spillover effects: LGAs with high non-compliance influenced neighboring LGAs. Table 5. Spatial lag model results. Variable Coefficient p-value Population density 0.011 0.003 Median income -0.007 0.018 Distance to major roads -0.018 0.004 Wetland presence 0.097 0.012 Permit approval rate -0.029 0.009 Political influence 0.046 0.020 Spatial lag (ρ) 0.26 0.000 3.4.3 Spatial Error Model (SEM) The SEM also performed well, but the SLM provided superior fit based on log-likelihood. In both models, population density, income, transport access, wetlands, and political influence remained significant drivers. 3.5 Robustness Checks • Multicollinearity: All VIF values < 3, indicating no serious multicollinearity. • Residual autocorrelation: Corrected under SLM specification. • Alternative weight matrices: Results consistent when using k-nearest neighbors (k=4). 3.6 Summary of Findings The analysis shows that land use change in Lagos is characterized by rapid urban expansion, with growth occurring largely at the expense of agricultural land and wetlands. This pattern reflects both the intensity of urbanization and the vulnerability of ecologically sensitive areas to conversion. Zoning non-compliance was found to be widespread, with more than 40 percent of new developments occurring outside officially designated zones. The problem is most acute in peri-urban LGAs such as Lekki and Badagry, where weak enforcement and speculative development pressures are especially pronounced. The spatial econometric analysis further revealed that population pressure, the availability of wetlands, proximity to major roads, low permit approval rates, and political influence are the strongest predictors of zoning non-compliance. These findings underscore the complex interplay of demographic, environmental, infrastructural, and institutional factors shaping land use outcomes in Lagos. 4. DISCUSSION 4.1 Principal Findings This study provides systematic evidence of the extent to which land use in Lagos has diverged from official zoning prescriptions, while also identifying the socio-economic, spatial, and institutional drivers of non-compliance. The results show that urban expansion between 2013 and 2023 occurred primarily at the expense of agricultural land and wetlands, with built-up areas increasing by nearly 43 percent. Over 40 percent of new developments were located in Volume-09 Issue 10, October-2025 ISSN: 2456-9348 Impact Factor: 8.232 International Journal of Engineering Technology Research Management (IJETRM) https://ijetrm.com/ IJETRM (http://ijetrm.com/) [98] zones inconsistent with planning allocations, with non-compliance particularly concentrated in peri-urban LGAs such as Eti-Osa, Ibeju-Lekki, and Badagry. Spatial econometric modeling revealed that population density, road accessibility, wetland presence, weak permit approval systems, and political influence were the strongest predictors of zoning violations, while spillover effects across LGAs suggested that non-compliance tends to spread geographically. Together, these findings highlight the deep misalignment between Lagos’ urban planning frameworks and its lived development patterns [31]. 4.2 Comparison with Existing Literature The results are consistent with prior research documenting Lagos’ rapid and largely unregulated urbanization. Studies using Landsat imagery have shown that built-up land in Lagos increased by more than 600 percent between the 1980s and the late 2010s, much of it occurring on wetlands and agricultural land [32,33]. Similarly, Akinmoladun and Adejumo identified widespread wetland conversion in the Lekki corridor, a finding echoed in this study’s hotspot analysis [34]. The prevalence of zoning non-compliance aligns with observations from other African cities, where zoning frameworks are often aspirational rather than enforceable. Kombe has described how weak institutions and rapid population growth create a “planning–implementation gap” in which regulatory frameworks have little bearing on urban outcomes [35]. Watson also emphasizes that African cities face a disconnect between urban visions and on-theground realities, a pattern strongly evident in Lagos [36]. The drivers of non-compliance identified here also mirror patterns reported elsewhere. Population growth has been shown to push informal housing expansion in both Lagos and Nairobi [37], while road access has been linked to speculative development in Dar es Salaam and Accra [38,39]. Wetland encroachment has emerged as a critical environmental issue in coastal African megacities, especially where land scarcity and real estate speculation intersect [40]. The role of political influence uncovered in this study resonates with Nigerian literature documenting elite-driven rezoning and selective enforcement of planning rules [41]. The negative relationship between permit approval rates and non-compliance suggests that strengthening institutional efficiency could serve as a corrective mechanism, echoing findings from Ghana and Uganda that effective permit systems reduce unauthorized construction [42,43]. 4.3 Implications for Urban Planning and Governance The evidence from Lagos raises important implications for planning practice. To begin with, the high rates of noncompliance challenge the relevance of traditional master planning. While official plans have long envisioned orderly, polycentric growth, actual development has been driven by informal practices, speculation, and institutional weakness. This reinforces Watson’s argument that many African cities face a fundamental disconnect between urban visions and lived realities [36]. Equally significant is the finding that 74 percent of non-compliant developments lacked permits, which underscores the limited enforcement capacity of Lagos’ planning institutions. Enforcement is not only hampered by inadequate technical staff but also undermined by political interference and corruption [44]. Addressing these institutional weaknesses requires a combination of increased funding, expanded technical training, and the deployment of geospatial technologies to improve monitoring. Another important dimension is the role of transport infrastructure, which emerges as a double-edged sword. Proximity to major roads was a significant predictor of non-compliance, suggesting that infrastructure expansion can inadvertently stimulate unauthorized development. This highlights the need to integrate transport investments with land use planning, ensuring that road corridors are pre-zoned and subject to strict monitoring [38,39]. Equally urgent are the environmental consequences of zoning violations. Wetland reclamation for residential estates and industrial facilities increases flood risks, undermines biodiversity, and reduces climate resilience. The catastrophic flooding in Lagos during 2022 illustrates the human and economic costs of neglecting ecological protection zones [45]. Stronger enforcement of conservation areas, supported by real-time monitoring, is therefore critical. A further implication is the detection of spillover effects, which underscores the need for metropolitan-scale governance. Since non-compliance in one LGA influences patterns in neighboring jurisdictions, fragmented local enforcement is unlikely to succeed. A metropolitan planning authority with cross-boundary jurisdiction could provide the necessary coordination for managing growth more coherently [46]. Volume-09 Issue 10, October-2025 ISSN: 2456-9348 Impact Factor: 8.232 International Journal of Engineering Technology Research Management (IJETRM) https://ijetrm.com/ IJETRM (http://ijetrm.com/) [99] 4.4 Policy Implications Several policy implications arise from these findings. Reforming zoning frameworks to reflect present-day realities is essential, particularly in rapidly urbanizing peri-urban areas. Adaptive and flexible zoning that accommodates mixeduse development can reduce pressures for non-compliance, provided ecological zones are strictly protected. Enforcement must be modernized through the use of drones, satellite monitoring, and automated change detection systems, while permit approval processes should be digitized and integrated with geospatial databases to improve transparency and reduce bureaucratic delays [47]. Transport and land use planning must be tightly coordinated, with infrastructure projects accompanied by proactive zoning enforcement to prevent opportunistic development. Institutional reforms are also needed to insulate planning agencies from political influence and corruption, including independent review boards and transparent amendment processes [41,44]. Protecting wetlands and conservation zones must remain a top priority, supported not only by legal sanctions but also by community-based stewardship programs [40,45]. Finally, participatory approaches to planning are vital. Involving residents in zoning decision-making enhances legitimacy and increases compliance, a lesson drawn from successful community-driven initiatives across Africa [36,42]. 4.6 Limitations While this study provides robust evidence, several limitations must be acknowledged. One important limitation relates to data gaps. Building permit records were incomplete, which constrained the ability to comprehensively distinguish between authorized and unauthorized developments. As a result, some informal or unpermitted activities may not have been fully captured in the analysis. Another limitation stems from classification uncertainty. Although the overall accuracy of land use classification exceeded 85 percent, there was occasional confusion between agricultural and wetland classes, particularly in areas were seasonal flooding blurred distinctions. This could have introduced minor errors in the detection of land use transitions. The treatment of zoning maps as static benchmarks also posed a challenge. In reality, zoning regulations in Lagos have been periodically revised, sometimes informally or without clear documentation. By relying on official zoning maps as fixed references, the study may not have fully accounted for such unrecorded adjustments, potentially affecting compliance assessments. A further limitation is the presence of unobserved variables. The econometric models did not explicitly capture factors such as land tenure complexity, community resistance to planning directives, or informal land market dynamics. Although difficult to quantify, these factors are highly influential in shaping compliance patterns. Finally, the temporal scope of the study was restricted to the period 2013–2023 due to the availability of consistent satellite imagery. Earlier periods of land use change were not examined, which limits the ability to analyze longerterm historical trajectories of compliance and non-compliance. 4.7 Directions for Future Research Future research should build on these findings while addressing the study’s limitations. One priority is the incorporation of dynamic zoning maps that account for temporal changes in regulations. Such an approach would provide a more nuanced understanding of compliance, especially in contexts where zoning rules are frequently revised, often informally. Similarly, the integration of cadastral data at the parcel level would enhance analysis by linking land use patterns more directly to tenure and ownership structures, thereby clarifying how property rights shape compliance behavior. Methodological innovation also offers promising avenues. Agent-based modeling, for example, could be used to simulate the behavior of developers, landowners, and regulators, capturing the micro-level decision-making processes that drive non-compliance. Comparative studies represent another important direction. Applying the same methodological framework to other African cities such as Nairobi, Accra, or Dar es Salaam would help to identify both commonalities and differences in compliance dynamics, enriching the broader discourse on urban governance. Finally, more research is needed on the social and environmental dimensions of zoning enforcement. Community engagement studies that explore how residents perceive zoning rules and enforcement practices could inform more participatory approaches to compliance. At the same time, integrating climate risk modeling with land use compliance