Search

The coastal regions of South Ecuador and Peru belong to the areas experiencing the strongest impact of the El Niño Southern Oscillation phenomenon. However, the impact and dynamic development of weather patterns during those events are not well understood, due to the sparse observational networks. In spite of neutral to cold conditions after the decaying 2015/16 El Niño in the central Pacific, the coastal region was hit by torrential rainfall in 2017 causing floods, erosion and landslides with many fatalities and significant damages to infrastructure. A new network of X-band weather radar systems in South Ecuador and North Peru allowed, for the first time, the spatio-temporally high-resolution monitoring of rainfall dynamics, also covering the 2017 event. Here, we compare this episode to the period 2014–2018 to point out the specific atmospheric process dynamics of this event. We found that isolated warming of the Niño 1 and 2 region sea surface temperature was the initial driver of the strong rainfall, but local weather patterns were modified by topography interacting with the synoptic situation. The high resolution radar data, for the first time, allowed to monitor previously unknown local spots of heavy rainfall during ENSO-related extreme events, associated with dynamic flow convergence initiated by low-level thermal breezes. Altogether, the coastal El Niño of 2017, at the same time, caused positive rainfall anomalies in the coastal plain and on the eastern slopes of the Andes, the latter normally associated only with La Niña events. Thus, the 2017 event must be attributed to the La Niña Modoki type.

Worldwide, machine learning (ML) is increasingly being used for developing flood early warning systems (FEWSs). However, previous studies have not focused on establishing a methodology for determining the most efficient ML technique. We assessed FEWSs with three river states, No-alert, Pre-alert and Alert for flooding, for lead times between 1 to 12 h using the most common ML techniques, such as multi-layer perceptron (MLP), logistic regression (LR), K-nearest neighbors (KNN), naive Bayes (NB), and random forest (RF). The Tomebamba catchment in the tropical Andes of Ecuador was selected as a case study. For all lead times, MLP models achieve the highest performance followed by LR, with f1-macro (log-loss) scores of 0.82 (0.09) and 0.46 (0.20) for the 1 h and 12 h cases, respectively. The ranking was highly variable for the remaining ML techniques. According to the g-mean, LR models correctly forecast and show more stability at all states, while the MLP models perform better in the Pre-alert and Alert states. The proposed methodology for selecting the optimal ML technique for a FEWS can be extrapolated to other case studies. Future efforts are recommended to enhance the input data representation and develop communication applications to boost the awareness of society of floods.

The Random Forest (RF) algorithm, a decision-tree-based technique, has become a promising approach for applications addressing runoff forecasting in remote areas. This machine learning approach can overcome the limitations of scarce spatio-temporal data and physical parameters needed for process-based hydrological models. However, the influence of RF hyperparameters is still uncertain and needs to be explored. Therefore, the aim of this study is to analyze the sensitivity of RF runoff forecasting models of varying lead time to the hyperparameters of the algorithm. For this, models were trained by using (a) default and (b) extensive hyperparameter combinations through a grid-search approach that allow reaching the optimal set. Model performances were assessed based on the R2, %Bias, and RMSE metrics. We found that: (i) The most influencing hyperparameter is the number of trees in the forest, however the combination of the depth of the tree and the number of features hyperparameters produced the highest variability-instability on the models. (ii) Hyperparameter optimization significantly improved model performance for higher lead times (12- and 24-h). For instance, the performance of the 12-h forecasting model under default RF hyperparameters improved to R2 = 0.41 after optimization (gain of 0.17). However, for short lead times (4-h) there was no significant model improvement (0.69 < R2 < 0.70). (iii) There is a range of values for each hyperparameter in which the performance of the model is not significantly affected but remains close to the optimal. Thus, a compromise between hyperparameter interactions (i.e., their values) can produce similar high model performances. Model improvements after optimization can be explained from a hydrological point of view, the generalization ability for lead times larger than the concentration time of the catchment tend to rely more on hyperparameterization than in what they can learn from the input data. This insight can help in th

Lack of rainfall information at high temporal resolution in areas with a complex topography as the Tropical Andes is one of the main obstacles to study its rainfall dynamics. Furthermore, rainfall types (e.g., stratiform, convective) are usually defined by using thresholds of some rainfall characteristics such as intensity and velocity. However, these thresholds highly depend on the local climate and the study area. In consequence, these thresholds are a constraining factor for the rainfall class definitions because they cannot be generalized. Thus, this study aims to analyze rainfall-event types by using a data-driven clustering approach based on the k-means algorithm that allows accounting for the similarities of rainfall characteristics of each rainfall type. It was carried out using three years of data retrieved from a vertically pointing Micro Rain Radar (MRR) and a laser disdrometer. The results show two main rainfall types (convective and stratiform) in the area which highly differ in their rainfall features. In addition, a mixed type was found as a subgroup of the stratiform type. The stratiform type was found more frequently throughout the year. Furthermore, rainfall events of short duration (less than 70 min) were prevalent in the study area. This study will contribute to analyze the rainfall formation processes and the vertical profile.

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.

© 2019 Elsevier B.V. Interest in atmosphere-surface flux modeling over the mountainous regions of the globe has increased recently, with a major focus on the prediction of water, carbon and other functional indicators in natural and disturbed conditions. However, less research has been centered on exploring energy fluxes (net radiation; sensible, latent and soil heat) and actual evapotranspiration (ETa) over the Neotropical Andean biome of the páramo. The present study assesses the implementation and parameterization of a state-of-art Land-Surface Model (LSM) for simulation of these fluxes over two representative páramo catchments of southern Ecuador. We evaluated the outputs of the LSM Community Land Model (CLM ver. 4.0) with (i) ground-level flux observations from the first (and highest) Eddy Covariance (EC) tower of the Northern Andean páramos; (ii) spatial ETa estimates from the energy balance-based model METRIC (based on Landsat imagery); and (iii) derived ETa from the closure of the water balance (WB). CLM's energy predictions revealed a significant underestimation on net radiation, which impacts the sensible and soil heat fluxes (underestimation), and delivers a slight overestimation on latent heat flux. Modeled CLM ETa showed acceptable goodness-of-fit (Pearson R = 0.82) comparable to ETa from METRIC (R = 0.83). Contrarily, a poor performance of ETa WB was observed (R = 0.46). These findings provide solid evidence on the CLM's accuracy for the ETa modeling, and give insights in the selection of other ETa methods. The study contributes to a better understanding of ecosystem functioning in terms of water loss through evaporative processes, and might help in the development of future LSMs’ implementations focused on climate / land use change scenarios for the páramo.

© 2019 by the authors. In remote areas with steep topography, such as the Tropical Andes, reliable precipitation data with a high temporal resolution are scarce. Therefore, studies focusing on the diurnal properties of precipitation are hampered. In this paper, we investigated two years of data from Micro-Rain Radars (MRR) in Cuenca, Ecuador, and Huaraz, Peru, from February 2017 to January 2019. This data allowed for a detailed study on the temporal precipitation characteristics, such as event occurrences and durations at these two locations. Our results showed that the majority of precipitation events had durations of less than 3 h. In Huaraz, precipitation has a distinct annual and diurnal cycle where precipitation in the rainy season occurred predominantly in the afternoon. These annual and diurnal cycles were less pronounced at the site in Cuenca, especially due to increased nocturnal precipitation events compared to Huaraz. Furthermore, we used a fuzzy logic classification of fall velocities and rainfall intensities to distinguish different precipitation types. This classification showed that nightly precipitation at both locations was predominantly stratiform, whereas (thermally induced) convection occurred almost exclusively during the daytime hours.

Weather radar networks are an excellent tool for quantitative precipitation estimation (QPE), due to their high resolution in space and time, particularly in remote mountain areas such as the Tropical Andes. Nevertheless, reduction of the temporal and spatial resolution might severely reduce the quality of QPE. Thus, the main objective of this study was to analyze the impact of spatial and temporal resolutions of radar data on the cumulative QPE. For this, data from the world’s highest X-band weather radar (4450 m a.s.l.), located in the Andes of Ecuador (Paute River basin), and from a rain gauge network were used. Different time resolutions (1, 5, 10, 15, 20, 30, and 60 min) and spatial resolutions (0.5, 0.25, and 0.1 km) were evaluated. An optical flow method was validated for 11 rainfall events (with different features) and applied to enhance the temporal resolution of radar data to 1-min intervals. The results show that 1-min temporal resolution images are able to capture rain event features in detail. The radar−rain gauge correlation decreases considerably when the time resolution increases (r from 0.69 to 0.31, time resolution from 1 to 60 min). No significant difference was found in the rain total volume (3%) calculated with the three spatial resolution data. A spatial resolution of 0.5 km on radar imagery is suitable to quantify rainfall in the Andes Mountains. This study improves knowledge on rainfall spatial distribution in the Ecuadorian Andes, and it will be the basis for future hydrometeorological studies.

Despite many efforts of the radar community, quantitative precipitation estimation (QPE) from weather radar data remains a challenging topic. The high resolution of X-band radar imagery in space and time comes with an intricate correction process of reflectivity. The steep and high mountain topography of the Andes enhances its complexity. This study aims to optimize the rainfall derivation of the highest X-band radar in the world (4450 m a.s.l.) by using a random forest (RF) model and single Plan Position Indicator (PPI) scans. The performance of the RF model was evaluated in comparison with the traditional step-wise approach by using both, the Marshall-Palmer and a site-specific Z−R relationship. Since rain gauge networks are frequently unevenly distributed and hardly available at real time in mountain regions, bias adjustment was neglected. Results showed an improvement in the step-wise approach by using the site-specific (instead of the Marshall-Palmer) Z−R relationship. However, both models highly underestimate the rainfall rate (correlation coefficient < 0.69; slope up to 12). Contrary, the RF model greatly outperformed the step-wise approach in all testing locations and on different rainfall events (correlation coefficient up to 0.83; slope = 1.04). The results are promising and unveil a different approach to overcome the high attenuation issues inherent to X-band radars.

Atmospheric carbon (CO 2) exchange, evapotranspiration (ET) processes, and their interactions with climatic drivers across tropical alpine grasslands are poorly understood. This lack of understanding is particularly evident for the páramo, the highest vegetated frontier in the northern Andes, the main source of water for inter-Andean cities, and a large carbon storage area. Studies of CO 2 and ET fluxes via the standard Eddy Covariance (EC) technique have never been applied to this region, limiting the understanding of diurnal / nocturnal exchanges and budget estimations. In this paper, we report the first EC analysis conducted on the Andean páramo (3765 m a.s.l.); this analysis measured CO 2, ET, and micrometeorological variables over two years (2016–2018) to understand their interactions with climatic / biophysical controls. The páramo was found to be a source of CO 2 and exhibited a net positive exchange …