Table 3 Synthesis of relationships among NDVI, EVI, and environmental variables in the Caatinga.
From: Predominantly positive XCO2 anomalies in the Caatinga biome highlight carbon vulnerability
Study/source | Vegetation types/region | Key relationships with NDVI | Key relationships with EVI | Relevant environmental controls | Notes |
|---|---|---|---|---|---|
This study (2025) | Caatinga (various phytophysiognomies) | Strong positive correlation with EVI (r > 0.86). Negative correlation with temperature. Positive correlation with precipitation | Strong positive correlation with NDVI (r > 0.86). Negative correlation with temperature. Positive correlation with precipitation | Hydrological seasonality; temperature-driven stress; structural differences among vegetation types | Explains spatial–temporal variability in XCO2 anomalies. |
Medeiros et al69. | Caatinga preserved vs. degraded areas | NDVI strongly driven by precipitation (r = 0.72–0.88). Sharp declines during droughts | EVI also tightly linked to rainfall and canopy structure; better captures degradation | Rainfall seasonality; drought intensity; land-use pressure | Shows different sensitivity between preserved and degraded areas. |
Zou et al70. | Caatinga dry forest sites | NDVI decreases with higher temperatures and VPD; increases with soil moisture and accumulated rainfall | Similar pattern to NDVI but more responsive to canopy density | Soil moisture; air temperature; vapor pressure deficit | Identifies temperature as a limiting factor during dry season. |
Barbosa et al71. | Shrublands and open Caatinga | NDVI highly seasonal; peaks depend on onset of rains. Weak vegetation activity under prolonged droughts | EVI more stable during early season moisture pulses | Hydrological pulses; early-rainfall events | Highlights differential phenological sensitivity. |
