Ecological Archives A025-078-A4

Katherine J. Willis, Alistair W. R. Seddon, Peter R. Long, Elizabeth S. Jeffers, Neil Caithness, Milo Thurston, Mathijs G. D. Smit, Randi Hagemann, and Marc Macias-Fauria. 2015. Remote assessment of locally important ecological features across landscapes: how representative of reality? Ecological Applications 25:1290–1302. http://dx.doi.org/10.1890/14-1431.1

Appendix D. Sensitivity analysis of discrete wavelet transform on the LEFT output of the Vulnerability layer (Table).

Table D1. Sensitivity analysis of discrete wavelet transform on the LEFT output of the Vulnerability layer. (Upper) Field forecast verification statistics with no offset (right), 50-cell offset (middle), and 100-cell offset (left). Verification statistics in Offset 100 marks the lower bound of verification statistics when accounting for the discretization artefact, suggesting a minor influence of this artefact on the agreement statistics between the two maps. (Lower) Example of changing the mother wavelet, applied to the highest resolution layer (1 × 1 cell, ~0.9 km²). Note that the verification statistics do not change with the selected mother wavelet, suggesting a robust result.

 

No Offset

 

Offset50

 

Offset100

 

Cell Size

Corr

MSE

Corr

MSE

Corr

MSE

16 × 16

0.62

0.012

0.55

0.011

0.55

0.012

8 × 8

0.66

0.010

0.57

0.011

0.56

0.012

4 × 4

0.65

0.010

0.59

0.01

0.56

0.012

2 × 2

0.62

0.012

0.58

0.011

0.55

0.013

1 × 1

0.61

0.012

0.55

0.012

0.55

0.013

1 × 1

Corr

MSE

Haar

0.61

0.012

Daubechies 4

0.61

0.012

Daubechies 6

0.61

0.012

Daubechies 10

0.61

0.012

Daubechies 20

0.58

0.013


[Back to A025-078]