Locating Old Trees Based On Remote Sensing Data

In dendroclimatological studies,old trees should be sampled in order to reconstruct longer past climate.It has always taken a lot of efforts for dendrochronologists to find old trees in the fields,especially where old trees were cut for different purposes.However,there are not continuous and high-resolution data of tree ages in a large spatial scale at present in China.Therefore,to find a way to quickly locate old tree groups is urgently needed in tree-ring and even ecological communities.In this paper a method is proposed to locate old tree groups that uses the relationship between remote sensing data and published tree age data.The method is based on the well demonstrated recognition that some indicator calculated from remote sensing data and stand ages often have a significant relationship.A Landsat 8 OLI(Operational Land Imager)131/38 remoting sensing image,acquired on 3 October 2016,is used to calculate 12 remote sensing indicators’ values.Generally,a tree-ring chronology is established using dozens of tree cores with missing rings and false rings being corrected for each core using the classical cross-dating technique.Therefore,the length of a chronology can represent the stand age well at the sampling site,and the lengths of the chronologies found in the study region are used as the stand age data.The stand age data are from 22 published tree-ring sampling sites which have specific tree-ring width chronologies’ lengths that have been used to establish regression models with the 12 remote sensing indicators using R language.Then,diagnose these models using statistics and R language.Finally,potential locations of old tree stands are indicated based on these models.The method proposed in this study could as an alternative to tree-age information,to help tree-ring researchers locate old tree stands quickly and accurately.