Details

Autor(en) / Beteiligte

• Godard, Jeanne
• Barriot, Jean-Pierre

Titel

Relationships between erosion rates and physiographic factors of drainage basins through a regression analysis: The case of Tahiti Nui Island

Ist Teil von

• Geomorphology (Amsterdam, Netherlands), 2022-05, Vol.404, p.108176, Article 108176

Ort / Verlag

Elsevier B.V

Erscheinungsjahr

2022

Quelle

Alma/SFX Local Collection

Beschreibungen/Notizen

• Tahiti Nui is a quiescent tropical volcanic island (French Polynesia) characterized by an intense erosion (up to 0.25 km3·kyr−1 over the last 1 Ma) that resulted in 27 main drainage basins radially distributed around a central depression. Those basins are characterized by different erosion rates (from 0.07 to 24.6 10−2 km3·kyr−1), ages (from 469 to 892 ka), micro-climate set-ups and other landform factors. In this study, we assess, in terms of explicative power, possible links between the erosion rates and relevant physiographic factors, through a regression analysis obeying the parsimony principle. The best regression models include, in decreasing order of importance, the following factors: the size (Planar Area or Maximal River Length) and/or the height (Maximal Altitude); the planar shape (Shape Factor or Relative Width); the inclination (Main Longitudinal Slope, Mean Slope, Maximum Altitude or Depth) and the mean precipitation (Average Annual Rainfall). The Erosion Duration and the positions along the North-South and East-West directions, which were used to highlight a possible effect of the winds, did not show significant effect on the erosion rates. In all the best fitting models, the greatest weight was associated with the size or the height factors, revealing that the highest erosion rates are the ones measured on the largest and/or highest basins. The best model is based on five physiographic factors per basin, in order of importance: Planar Area, Relative Width, Maximum Altitude, Main Longitudinal Slope and Average Annual Rainfall. The Planar Area factor seems to be the most relevant parameter since it explains by itself the largest part of the erosion rates. Finally, the role of average annual rainfall is not consistent between the regression models. This suggests that a more suitable physiographic factor to constrain the effect of precipitations on erosion rates must be considered, probably linked to the temporal climate variability as well as the frequency of cyclone-driven rainfall. Besides, the weight of the Average Annual Rainfall factor was generally positive and small in most regression models, synonymous of a minor role. •The two best models of basin erosion are based on four or five physiographic factors.•Long-term erosion rates are highly correlated with the planar areas of the basins.•Average annual rainfall has a low impact on erosion rates.•Erosion durations do not influence erosion rates.