ObjectiveThe normalized difference vegetation index (NDVI) is one of the most widely used remote sensing data and has been widely used in the estimation of the net primary productivity (NPP). But the shortcoming of NDVI is that the actual measured data are relatively short. Reconstruction of NDVI sequences based on tree-ring data can compensate for the shortcomings of NDVI time series to some extent.
MethodThis study analyzed the potential relationship between NDVI and tree ring width data of Pinus koraiensis, Abies nephrolepis and Quercus mongolica in the Danqinghe Forest Farm in the southern Xiaoxing'an Mountains of northeastern China, and attempted to reconstruct the past NDVI changes in Xiaoxing'an Mountains.
ResultThe results showed that the standard chronologies of Pinus koraiensis, Abies nephrolepis and Quercus mongolica were significantly positively correlated with NDVI in July of current year. The correlation between the synthesized chronology of multiple species in the plot and the NDVI in July was higher than that of each single species. The synthesized chronology of multiple species and the NDVI in July have a consistent correlation with climatic factors (temperature and precipitation), they are positively correlated with temperature from October of the previous year to March of the growing year and negatively correlated with temperature from April to July of growing year. Therefore, we reconstructed the past 115 years NDVI in July according to the relationship between synthesized chronology and NDVI in July and using the linear regression model to fit the equation. The linear regression model can explain a variance amounted to 42.6% (Radj2=0.401, F=17.08, P < 0.000 1). The NDVI reconstruction showed that there were two high-value periods (1967-1978, 1986-1995) and three low-value periods (1920-1925, 1949-1954, 1999-2004) since 1900. The low value period of NDVI coincided with the historical record of drought, so drought may be the main cause of low NDVI values. In addition, factors such as deforestation and forest fire may also lead to the decrease of NDVI value.
ConclusionIt is worth noting that the NDVI projection range and multi-species combinations should be considered to reduce reconstruction errors when establishing the relationship between tree rings and NDVI.