TY - JOUR
T1 - Spatiotemporal analysis of forest cover change and associated environmental challenges
T2 - a case study in the Central Highlands of Vietnam
AU - Tran, Duy X.
AU - Tran, Thuong V.
AU - Pearson, Diane
AU - Myint, Soe W.
AU - Lowry, John
AU - Nguyen, Tuan T.
N1 - Publisher Copyright:
© 2021 Informa UK Limited, trading as Taylor & Francis Group.
PY - 2021
Y1 - 2021
N2 - Spatiotemporal regression combining Theil-Sen median trend and Man-Kendall tests was applied to MODIS time-series data to quantify the trend and rate of change to forest cover in the Central Highlands, Vietnam from 2001 to 2019. Several MODIS data products, including Percent Tree Cover (PTC), Evapotranspiration (ET), Land Surface Temperature (LST), and Gross Primary Productivity (GPP) were selected as indicators for forest cover and climate and carbon cycle patterns. Emerging hot spot analysis was applied to identify patterns of long-term deforestation. Spatial regression analysis using Geographically Weighted Regression (GWR) was performed to understand variations in the relationship between vegetation changes and trends in LST, ET, and GPP. Our analysis reveals that deforestation occurred significantly in the study area with a total decrease of 14.5% in PTC and a total of 7314 deforestation hot spots were identified. Results indicate that forest cover loss explains 72.9%, 67.7%, and 89.4% of the changes in ET, GPP, and LST, respectively, and the levels of influence are heterogenous across space and dependent on the types of deforestation hot spots. The approach introduced in our study can be performed worldwide to address complex research questions about environmental challenges that emerge from deforestation.
AB - Spatiotemporal regression combining Theil-Sen median trend and Man-Kendall tests was applied to MODIS time-series data to quantify the trend and rate of change to forest cover in the Central Highlands, Vietnam from 2001 to 2019. Several MODIS data products, including Percent Tree Cover (PTC), Evapotranspiration (ET), Land Surface Temperature (LST), and Gross Primary Productivity (GPP) were selected as indicators for forest cover and climate and carbon cycle patterns. Emerging hot spot analysis was applied to identify patterns of long-term deforestation. Spatial regression analysis using Geographically Weighted Regression (GWR) was performed to understand variations in the relationship between vegetation changes and trends in LST, ET, and GPP. Our analysis reveals that deforestation occurred significantly in the study area with a total decrease of 14.5% in PTC and a total of 7314 deforestation hot spots were identified. Results indicate that forest cover loss explains 72.9%, 67.7%, and 89.4% of the changes in ET, GPP, and LST, respectively, and the levels of influence are heterogenous across space and dependent on the types of deforestation hot spots. The approach introduced in our study can be performed worldwide to address complex research questions about environmental challenges that emerge from deforestation.
KW - Deforestation
KW - MODIS
KW - environmental degradation
KW - environmental management
KW - spatiotemporal regression
UR - http://www.scopus.com/inward/record.url?scp=85121496056&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85121496056&partnerID=8YFLogxK
U2 - 10.1080/10106049.2021.2017013
DO - 10.1080/10106049.2021.2017013
M3 - Article
AN - SCOPUS:85121496056
JO - Geocarto International
JF - Geocarto International
SN - 1010-6049
ER -