Assessment of Native Radar Reflectivity and Radar Rainfall Estimates for Discharge Forecasting in Mountain Catchments with a Random Forest Model

Johanna Orellana-Alvear,Rütger Rollenbeck,Rolando Célleri,Pablo Contreras,Paul Muñoz,Jörg Bendix

doi:10.3390/rs12121986

Johanna Orellana-Alvear, Rütger Rollenbeck + Show 4 more

Open Access

https://doi.org/10.3390/rs12121986

Copy DOI

Abstract

Discharge forecasting is a key component for early warning systems and extremely useful for decision makers. Forecasting models require accurate rainfall estimations of high spatial resolution and other geomorphological characteristics of the catchment, which are rarely available in remote mountain regions such as the Andean highlands. While radar data is available in some mountain areas, the absence of a well distributed rain gauge network makes it hard to obtain accurate rainfall maps. Thus, this study explored a Random Forest model and its ability to leverage native radar data (i.e., reflectivity) by providing a simplified but efficient discharge forecasting model for a representative mountain catchment in the southern Andes of Ecuador. This model was compared with another that used as input derived radar rainfall (i.e., rainfall depth), obtained after the transformation from reflectivity to rainfall rate by using a local Z-R relation and a rain gauge-based bias adjustment. In addition, the influence of a soil moisture proxy was evaluated. Radar and runoff data from April 2015 to June 2017 were used. Results showed that (i) model performance was similar by using either native or derived radar data as inputs (0.66 < NSE < 0.75; 0.72 < KGE < 0.78). Thus, exhaustive pre-processing for obtaining radar rainfall estimates can be avoided for discharge forecasting. (ii) Soil moisture representation as input of the model did not significantly improve model performance (i.e., NSE increased from 0.66 to 0.68). Finally, this native radar data-based model constitutes a promising alternative for discharge forecasting in remote mountain regions where ground monitoring is scarce and hardly available.

Highlights

Discharge forecasting is of main importance for water management and decision-making support all around the globe
Several studies [1,3] have proven their efficiency and good performance at different catchment scales and often at daily or monthly lead times. These models often come with high-computational costs derived from the increase of the spatial resolution of relevant variables used in the model as in Heuvelink et al [6]
Random Forest (RF) is a decision tree-based model from the machine learning family. It is an ensemble method, which means that several trees are built and the predicted estimation is the combination of the results of all tree models in the forest

Summary

Introduction

Discharge forecasting is of main importance for water management and decision-making support all around the globe. Several studies [1,3] have proven their efficiency and good performance at different catchment scales and often at daily or monthly lead times These models often come with high-computational costs derived from the increase of the spatial resolution of relevant variables used in the model as in Heuvelink et al [6]. Remote and mountainous regions are usually scarcely monitored which restrict the use of distributed models for discharge forecasting. This is due to the necessity of spatially detailed description of several hydro-geomorphological variables on the study area, which are frequently limited or non-existing. The authors found that the location and number of rain gauge highly influenced the model uncertainty

Objectives

Methods

Results

Conclusion