*4.3. Correlation between Climatic Factors and NDVI*

Consistent with previous research, the correlation coefficient between NDVI and precipitation along the GH highway showed an insignificant negative correlation. During the nearly 30 years of the operation period, the slight decrease in total precipitation did not cause major a catastrophe for vegetation growth in the Southwest China; instead, the change in precipitation frequency made local rain recruitment more frequent, which partly compensated for the growth of southwestern vegetation being limited by the alternating time of dry and wet periods, rather than total precipitation [51]. At the same time, the partial correlation coefficients of NDVI with precipitation and temperature were 0.05 and 0.17, respectively; both were less than 0.5, which is basically consistent with previous research [29]. The proportion of karst landform in the study area is high (76.5%); together with the large change of karst underlying surface and the high degree of topographic relief, the spatial heterogeneity of temperature within a small range may be high. This results in temperature being a major factor controlling vegetative growth over a small range. A previous study has shown that terrain is generally a covariate of temperature, which is highly consistent with temperature change. Therefore, we did not consider terrain factors as covariates to participating in the partial correlation analysis [29].

Similar to previous studies, the correlation between NDVI and precipitation was weakly positive on the artificial surface, while the correlation between NDVI and temperature was low and negative on artificial surfaces [30]. Human activities are strong in artificial surface areas, and the environment is more fragile than in other areas. The low precipitation infiltration of hardened surface exacerbates water shortages and temperature increases, which may force the growth of vegetation to be slow. The surrounding ecology is fragile. The increase of precipitation makes the water supply needed for the growth of regional vegetation sufficient, thus reducing the vulnerability, while artificial surface vegetation is more sensitive to drought.

#### **5. Conclusions**

In this paper, NDVI and climate data were used to analyze the influence of the GH highway on the area within 8 km of the route. It was found that the annual mean and growth rate of NDVI in the core area within the construction and operation periods of the GH highway were smaller than those in the contrast area, the inter-annual variation fluctuated greatly, and the influence on the area was mainly within 2 km of the GH highway. Within the operation period, the NDVI reached a peak and then decreased slightly. Within the construction and operation period, the NDVI along the route increased overall, precipitation showed a downward trend, and temperature showed an upward trend. The correlation between NDVI and climate factors indicated that the correlation between NDVI and temperature is stronger than that between NDVI and precipitation. The influence of LUCC on NDVI was mainly manifested as an increase in artificial cover surface and the decline of other land-use types, resulting in the change of NDVI.

#### **6. Limitations and Prospects**

In this study, only the inter-annual variability of climate variables in response to NDVI was considered, seasonal variation of climate indicators was not considered; furthermore, only the annual maximum value of Landsat was used to synthesize the NDVI images, and multi-source remote sensing data fusion methods were not considered. In the future, multi-source remote sensing can be used to explore the corresponding relationship between NDVI and other climatic factors (e.g., surface soil humidity, evaporation, seasonal drought, and so on). Remote sensing data under nighttime lighting can also be adopted, in order to explore the correlation between vegetation cover and human activities.

**Author Contributions:** Conceptualization, Y.W. and S.L.; methodology, G.L.; software, L.G. and J.Y.; formal analysis, C.G. and F.Y.; investigation, L.G., J.Y., Z.S. and X.Y.; data curation, Y.C., H.P. and X.Y.; writing—original draft preparation, Y.W.; writing—review and editing, C.G. and S.L.; visualization, L.G.; supervision, Y.X.; project administration, G.L.; funding acquisition, Y.W. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by Guizhou Provincial Science and Technology Projects (ZK [2022] YB334) and Doctoral Program of Guizhou Education University (X2021049).

**Data Availability Statement:** Not applicable.

**Acknowledgments:** We thank the anonymous reviewers for their valuable comments. We gratefully acknowledge the design of S.L. and the contribution of co-authors.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**

