Detecting Long-term Changes in Vegetation Condition using NDVI time-series and Non-parametric Trend Analysis in Calabar Metropolis

Abstract

Long-term monitoring of vegetation dynamics is essential for understanding ecosystem responses to climate variability and land-use change, particularly in rapidly urbanizing regions. This study assessed decadal vegetation greenness in Calabar Metropolis, Nigeria, using satellite-derived Normalized Difference Vegetation Index (NDVI) data spanning 2015–2025. NDVI time series were derived from Landsat 7, Landsat 8, and Landsat 9 imagery and analyzed using descriptive statistics, Mann–Kendall trend analysis, Sen’s slope estimation, Pettitt’s test, segmented regression, and anomaly analysis. NDVI values ranged from 0.10 to 0.61, with an overall mean of 0.39 and median of 0.38. Annual mean NDVI varied from 0.303 ± 0.084 in 2016 to 0.468 ± 0.074 in 2025. The Mann–Kendall test revealed a significant increasing trend in annual NDVI (τ = 0.491, Z = 2.024, p = 0.043). Sen’s slope indicated a positive rate of change of 0.0098 NDVI units year⁻¹ (95% CI: 0.0004–0.0179). The coefficient of variation declined from 27.5% in 2016 to 15.8% in 2025, indicating increasing vegetation stability. Pettitt’s test identified 2018 as the most likely change point, but this shift was not significant (U* = 22, p = 0.271). Monthly mean DVI was lowest in February (0.302 ± 0.057) and highest in November (0.553 ± 0.025). Standardized anomalies ranged from z = −3.03 in March 2018 to z = 2.35 in May 2022, reflecting episodic stress and rapid recovery. The results indicate a gradual and statistically significant increase in vegetation greenness over the 2015–2025 period, with no evidence of abrupt structural change.