Table 2 Landslide detection methods based on empirical rules compared in this study

1

Shimada et al. (2014), Watanabe et al. (2016)*

\(\gamma_{{{\text{HH}},{\text{VV}}}} > 0.65\)

2

Shimada et al. (2014), Watanabe et al. (2016)*

\(\bar{\alpha } < 35^\circ\)

3

Watanabe et al. (2012)*

\(P_{\text{V}} < 45\% {\text{AND }}P_{\text{D}} < 10\%\)

\({\text{AND }}P_{\text{S}} > 10\%\)

4

Shibayama et al. (2015)

\(P_{\text{S}} > 60\% \;({\text{if}} \,\,LIA < 30^\circ )\)

\(P_{\text{S}} > 40\% {\text{ AND }}P_{\text{S}} > P_{\text{V}} \;({\text{if }} 30^\circ < LIA)\)

5

This study (PolSAR)

\(T_{11} > 60\% \;({\text{if }}LIA < 20^\circ )\)

\(T_{11} > 50\% \;({\text{if}} \,20^\circ < LIA < 30^\circ )\)

\(T_{11} > 40\% {\text{ AND }}T_{11} > T_{33} \;({\text{if }} 30^\circ < LIA)\)

6

This study (PolSAR + InSAR)

\(T_{11} > 60\% {\text{ AND }}V_{\theta } > 0.05 \;({\text{if }}LIA < 20^\circ )\)

\(T_{11} > 50\% {\text{ AND }}V_{\theta } > 0.05 \;({\text{if }} 20^\circ < LIA < 30^\circ )\)

\(T_{11} > 40\% {\text{ AND }}T_{11} > T_{33} {\text{ AND }}V_{\theta } > 0.05 \;({\text{if }} 30^\circ < LIA)\)