Stochastic assessment of landslides and debris flows in the Jemma basin, Blue Nile, Central Ethiopia

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216208%3A11310%2F16%3A10328713" target="_blank" >RIV/00216208:11310/16:10328713 - isvavai.cz</a>

Výsledek na webu

<a href="http://dx.doi.org/10.4461/GFDQ.2014.395" target="_blank" >http://dx.doi.org/10.4461/GFDQ.2014.395</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.4461/GFDQ.2014.395" target="_blank" >10.4461/GFDQ.2014.395</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Stochastic assessment of landslides and debris flows in the Jemma basin, Blue Nile, Central Ethiopia

Popis výsledku v původním jazyce

In this paper we evaluate a stochastic method to assess the spatial distribution of landslide and debris flow processes in the Jemma basin, Central Ethiopia. The Jemma basin is draining the highlands (max. 3,676 m a.s.l.) northeast of Addis Ababa towards the Blue Nile. The basin is characterized by a deeply incised stratigraphy made up of volcanic deposits like flood basalts and tephra. Hence, gravitational mass movements as well as water driven erosion processes occur, documented by the respective forms. We mapped these features using Google Earth images, aerial photo interpretation and fieldwork. The information about the spatial distribution of landslide and debris flow forms was taken as dependent variable in the stochastic modelling approach. Moreover, we performed a detailed terrain analysis to derive the independent variables. We applied two different stochastic modelling approaches based on i) boosted regression trees (BRT) and ii) on an maximum entropy method (MEM) to predict the potential spatial distribution of landslide and debris flows in the Jemma basin. The models are statistically evaluated using the training data and a set of performance parameters such as the area under the receiver operating characteristic curve (AUC). Variable importance and response curves provide further insight into controlling factors of landslide and debris flow distribution. The study shows that both processes can be perfectly identified and distinguished. The spatial distribution of the predicted process susceptibilities generally follows topographic constraints. Model performance parameters show better results for BRT, that outperforms MEM. However, MEM results are quite robust and hence are used for the spatial prediction of process susceptibilities.

Název v anglickém jazyce

Stochastic assessment of landslides and debris flows in the Jemma basin, Blue Nile, Central Ethiopia

Popis výsledku anglicky

In this paper we evaluate a stochastic method to assess the spatial distribution of landslide and debris flow processes in the Jemma basin, Central Ethiopia. The Jemma basin is draining the highlands (max. 3,676 m a.s.l.) northeast of Addis Ababa towards the Blue Nile. The basin is characterized by a deeply incised stratigraphy made up of volcanic deposits like flood basalts and tephra. Hence, gravitational mass movements as well as water driven erosion processes occur, documented by the respective forms. We mapped these features using Google Earth images, aerial photo interpretation and fieldwork. The information about the spatial distribution of landslide and debris flow forms was taken as dependent variable in the stochastic modelling approach. Moreover, we performed a detailed terrain analysis to derive the independent variables. We applied two different stochastic modelling approaches based on i) boosted regression trees (BRT) and ii) on an maximum entropy method (MEM) to predict the potential spatial distribution of landslide and debris flows in the Jemma basin. The models are statistically evaluated using the training data and a set of performance parameters such as the area under the receiver operating characteristic curve (AUC). Variable importance and response curves provide further insight into controlling factors of landslide and debris flow distribution. The study shows that both processes can be perfectly identified and distinguished. The spatial distribution of the predicted process susceptibilities generally follows topographic constraints. Model performance parameters show better results for BRT, that outperforms MEM. However, MEM results are quite robust and hence are used for the spatial prediction of process susceptibilities.

Druh

J_x - Nezařazeno - Článek v odborném periodiku (Jimp, Jsc a Jost)

CEP obor

DE - Zemský magnetismus, geodesie, geografie

OECD FORD obor

—

Projekt

<a href="/cs/project/GCP209%2F12%2FJ068" target="_blank" >GCP209/12/J068: Výzkum svahových pohybů a eroze jako indikátoru morfotektonické aktivity Etiopské vysočiny založený na DPZ</a><br>

Návaznosti

P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)

Rok uplatnění

2016

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název periodika

Geografia Fisica e Dinamica Quaternaria

ISSN

0391-9838

e-ISSN

—

Svazek periodika

39

Číslo periodika v rámci svazku

1

Stát vydavatele periodika

IT - Italská republika

Počet stran výsledku

8

Strana od-do

51-58

Kód UT WoS článku

000389169500005

EID výsledku v databázi Scopus

2-s2.0-84999176904