Temporal and Spatial Changes and Driving Forces of NDVI From 1982 - 2015 in Qinba Mountains, China

: The spatiotemporal variation and driving force of Normalized Difference 11 Vegetation Index (NDVI) is helpful to regional ecological environment protection and 12 natural resource management. Using the Sen and Mann–Kendall methods, Hurt index, 13 Space transfer matrix and Geodetector, this study investigated the temporal and 14 spatial changes and driving forces of NDVI during 1982 - 2015. The results showed 15 that: （ 1 ） For the period 1982 to 2015, the high vegetation coverage was mainly 16 distributed in Qinling Mountains and Daba mountain, while the value of NDVI was 17 low in high altitude area in the west, low altitude in the East and Hanjiang River 18 valley. （ 2 ） The change trend of NDVI in Qinba Mountains is mainly to maintain stable 19 and slow growth. And the slow growth changes significantly. NDVI increased slowly 20 mainly in the East and northwest. （ 3 ） The future change trend of NDVI in Qinba 21 Mountain is mainly slow growth and stability, which indicates that the ecological 22 construction in Qinba Mountains is good. (4) Through the geographical detector, the 23 main factors affecting NDVI in Qinba Mountains are natural factors mainly including rainfall, soil type and digital elevation model (DEM), while human activities mainly 25 including population density have little influence on NDVI in Qinba Mountains. 26 Natural environment factors and human activities make a great difference on the 27 spatial distribution of NDVI. This study provides a help for the sustainable 28 development of the naturel environment in Qinba Mountains. Abstract : The spatiotemporal variation and driving force of Normalized Difference 64 Vegetation Index (NDVI) is helpful to regional ecological environment protection and 65 natural resource management. Using the Sen and Mann–Kendall methods, Hurt index, 66 Space transfer matrix and Geodetector, this study investigated the temporal and 67 spatial changes and driving forces of NDVI during 1982 - 2015. The results showed 68 that: （ 1 ） For the period 1982 to 2015, the high vegetation coverage was mainly 69 distributed in Qinling Mountains and Daba mountain, while the value of NDVI was 70 low in high altitude area in the west, low altitude in the East and Hanjiang River 71 valley. （ 2 ） The change trend of NDVI in Qinba Mountains is mainly to maintain stable 72 and slow growth. And the slow growth changes significantly. NDVI increased slowly 73 mainly in the East and northwest. （ 3 ） The future change trend of NDVI in Qinba 74 Mountain is mainly slow growth and stability, which indicates that the ecological 75 construction in Qinba Mountains is good. (4) Through the geographical detector, the Natural environment factors and human activities make a great difference on the spatial distribution of NDVI. This study provides a help for the sustainable 81 development of the naturel environment in Qinba Mountains. 82


Temporal and spatial changes and driving forces of NDVI
Meanwhile, vegetation is also most sensitive to environmental changes, especially 89 climate changes (Wen et al. 2017). Thus, it is meaningful for the ecological 90 environment to study the vegetation change and its driving forces . 91 As one of the most popular vegetation indices used for monitoring short-term 92 and long-term variations of vegetation (Jiang et al. 2016), NDVI has been extensively 93 used to show the level of regional vegetation coverage and vegetation growth status, 94 which has been applied in many areas of research, such as vegetation spatial 95 distribution, vegetation dynamic monitoring and dynamic evolution (Cao et al. 2008). 96 Therefore, it can not only describe the temporal and spatial changes of vegetation, but 97 also reflects the feedback of climate (Miao et al. 2012).    This dataset is suitable for long-term monitoring of NDVI because of it is the most 150 extensive data in time range, and after processing and correction, the data accuracy Where n is the number of observations. xj and xk are the ranks of observations xi and 201 xj of the time series. When the MK trend has a significance level greater than 5% (p < 202 0.05, ZMK ≥ |±1.96|), the NDVI trend is considered significant (Hamed 2009).
2．To calculate the cumulative deviation, It is used to compare the effects of two factors X1 and X2 on the spatial 252 distribution of attribute y. If there is a significant difference, it is recorded as "Y", 253 otherwise it is recorded as "n". According to this, we can judge which factor has more 254 influence on NDVI.      Table 2.      other factors, and the action mechanism of the dominant natural factors was different.

393
The impact of land use type, aspect and GDP on the spatial distribution of NDVI 394 has significant difference, but has no significant difference with other factors.

395
Population density, slope, aspect and soil type had significant differences on the 396 spatial distribution of NDVI, but had no significant difference with other factors.
397 Table 5.Statistical significance of the detection factors for NDVI distribution 398 Using F test with significance level of 0.05, y means that there is significant difference between the two factors in 399 the spatial distribution of NDVI; n means that there is no significant difference 400 3.5.3Interaction detector 401 Table 6 shows the results of interaction detector, The detection results show that 417  The geographical position of Qinba Mountains Note: The designations employed and the presentation of the material on this map do not imply the expression of any opinion whatsoever on the part of Research Square concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. This map has been provided by the authors.

Figure 2
Spatial pattern of vegetation cover Note: The designations employed and the presentation of the material on this map do not imply the expression of any opinion whatsoever on the part of Research Square concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. This map has been provided by the authors.

Figure 3
Area ratio of NDVI Figure 4 The variation tendency of annual NDVI in 1982 to 2015 Note: The designations employed and the presentation of the material on this map do not imply the expression of any opinion whatsoever on the part of Research Square concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. This map has been provided by the authors.

Figure 5
The signi cance of variation tendency of annual NDVI in 1982 to 2015 Note: The designations employed and the presentation of the material on this map do not imply the expression of any opinion whatsoever on the part of Research Square concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. This map has been provided by the authors.

Figure 6
The variation tendency of vegetation in future Note: The designations employed and the presentation of the material on this map do not imply the expression of any opinion whatsoever on the part of Research Square concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. This map has been provided by the authors.