Skip to main content

Table 1 Conditioning factors in detail

From: Stepwise integration of analytical hierarchy process with machine learning algorithms for landslide, gully erosion and flash flood susceptibility mapping over the North-Moungo perimeter, Cameroon

Factor (unit)

Basic expression

Type

Scale (m): original /resampled

Elevation (m)

N/A

Continuous

30 / 20

Slope (angle) (°)

\(arctan\cdot \sqrt{{f}_{x}^{2}+{f}_{y}^{2}}\)

Continuous

30 / 20

Slope (aspect) (°)

\({270}^{0}+arctan\cdot \left(\frac{{f}_{y}}{{f}_{x}}\right)-90^\circ \frac{{f}_{x}}{\left|{f}_{y}\right|}\)

Continuous

30 / 20

Plan curvature (m/100)

Convex value > 0; concave value < 0; Flat elsewhere

Continuous

30 / 20

Profile curvature (m/100)

-//-

Continuous

30 / 20

Relative slope position

Thresholding of Topographic Position Index (TPI), expressed as:

\(\frac{{E}_{Pixel}}{{E}_{Surrounding}}\)

where \({E}_{Pixel}\) is the elevation of the cell, and \({E}_{Surrounding}\) is the mean elevation of the neighbor pixels

Categorical

30 / 20

TRI

\(\sqrt{\left|X\right|\left({max}^{2}-{min}^{2}\right)}\)

where \(X\) is the elevation of each neighbor cell to a specific cell (0,0) (m), and \(max\) and \(min\) are the largest and smallest elevations among the nine neighboring pixels

Continuous

30 / 20

LS

\({\left(\frac{{A}_{s}}{22.13}\right)}^{m}\cdot {\left(\frac{\mathrm{sin}(\theta )}{0.0896}\right)}^{n}\)

As is the unit contributing area (m), θ is the slope in radians, and m (0.4–0.56) and n (1.2–1.3) are exponents

Continuous

30 / 20

Valley depth (m)

N/A

Continuous

30 / 20

Lithology

N/A

Categorical

1:5 00 000 / 20

Fault density

\(\left[ {\left( {L_{1} \cdot V_{1} } \right) + \left( {L_{1} \cdot V_{1} } \right)} \right]/ \oplus\), where:

\(\oplus =\) radius;

\(L=\) portion of line within \(\oplus\);

\(V=\) field value

Continuous

30 / 20

Distance to fault (m)

Euclidean distance (two dimensions):

\(\Delta =\sqrt{\left[({{x}_{2}-{x}_{1})}^{2}+\right]({{y}_{2}-{y}_{1})}^{2}},\) where:

∆ = Euclidean distance;

\({(x}_{2}-{x}_{1})\)= coordinate of 1rst point, i.e., on one side of the fault;

\({(y}_{2}-{y}_{1})\)= coordinate of 2nd point, i.e., on other side of the fault

Continuous

30 / 20

Soil depths (cm)

USDA global model

Continuous

250 / 20

Soil types

N/A

Categorical

1:1 000 000 / 20

Drainage density

Same as Fault density

Continuous

30 / 20

Distance to stream (m)

Same as Distance to fault

Continuous

30 / 20

Channel network distance (m)

Same as Distance to fault

Continuous

30 / 20

SPI

\({A}_{s}*tan\beta\)

Continuous

30 / 20

TWI

\({Log}_{e}\left(\frac{{A}_{s}}{tan\beta }\right)\)

Continuous

30 / 20

Mean annual rainfall (MAR)- IDW or (MFI)

\({Z}_{P}=\frac{\sum_{i=1}^{n}\left(\frac{{z}_{i}}{{d}_{i}^{p}}\right)}{\sum_{i=1}^{n}\left(\frac{1}{{d}_{i}^{p}}\right)}\) Or \(MFI={\sum }_{i=1}^{i=12}\frac{{P}_{1}^{2}}{P}\)

Continuous

4 000 / 20

Distance to road (m)

Same as Distance to fault

Continuous

20

Road density

Same as Fault Density

Continuous

20

LULC

Random Forest

Categorical

20

SEVI

\(\frac{NIR}{Red}*\left(\frac{{SWIR1}_{mean}-{SWIR1}_{min}}{{SWIR1}_{max}-{SWIR1}_{min}}\right)+0.581*\frac{1}{Red}\)

Continuous

20