Changes in pollution pressure exerted on the Rocha River in Bolivia from diffuse sources were assessed using potential non-point pollution indexes (PNPI) for 1997 and 2017. PNPI is a simple, low-effort, time- and resource-saving method suitable for data-scarce regions, as it works at catchment level with commonly available geographical data. Land use type (obtained by Landsat imagery classification), runoff (determined by runoff coefficient characterisation) and distance to river network (calculated at perpendicular distance) were each transformed into corresponding indicators to determine their relative importance in generating pollution. Weighted sum, a multi-criteria analysis tool in the GIS environment, was used to combine indicators with weighting values. Different weighting values were assigned to each of the indicators resulting in a set of six equations. The results showed that higher PNPI values corresponded to human settlements with high population density, higher runoff values and shorter distance to river network, while lower PNPI values corresponded to semi-natural land use type, lower runoff coefficient and longer distances to river. PNPI values were positively correlated with measured nitrate and phosphate concentrations at six sub-catchment outlets. The correlation was statistical significant for phosphate in 2017. Maps were produced to identify priority source areas that are more likely to generate pollution, which is important information for future management.

The transition to a more environmentally friendly energy matrix by reducing fossil fuel usage has become one of the most important goals to control climate change. Variable renewable energy sources (VRES) are a central low-carbon alternative. Nevertheless, their variability and low predictability can negatively affect the operation of power systems. On this issue, energy-system-modeling tools have played a fundamental role. When exploring the behavior of the power system against different levels of VRES penetration through them, it is possible to determine certain operational and planning strategies to balance the variations, reduce the operational uncertainty, and increase the supply reliability. In many developing countries, the lack of such proper tools accounting for these effects hinders the deployment potential of VRES. This paper presents a particular energy system model focused on the case of Bolivia. The model manages a database gathered with the relevant parameters of the Bolivian power system currently in operation and those in a portfolio scheduled until 2025. From this database, what-if scenarios are constructed allowing us to expose the Bolivian power system to a set of alternatives regarding VRES penetration and Hydro storage for that same year. The scope is to quantify the VRES integration potential and therefore the capacity of the country to leapfrog to a cleaner and more cost-effective energy system. To that aim, the unit-commitment and dispatch optimization problem are tackled through a Mixed Integer Linear Program (MILP) that solves the cost objective function within its constraints through the branch-and-cut method for each scenario. The results are evaluated and compared in terms of energy balancing, transmission grid capability, curtailment, thermal generation displacement, hydro storage contribution, and energy generation cost. In the results, it was found that the proposed system can reduce the average electricity cost down to 0.22 EUR/MWh and also reduce up to 2.22 × 10${}^6$ t (96%) of the CO${}_2$ emissions by 2025 with very high penetration of VRES but at the expense of significant amount of curtailment. This is achieved by increasing the VRES installed capacity to 10,142 MW. As a consequence, up to 7.07 TWh (97%) of thermal generation is displaced with up to 8.84 TWh (75%) of load covered by VRES.

Climate change and variability are likely to increase in most parts of the world, leading to more extreme events, which may increase the stress on already threatened water resources. This study focuses on the effects of the El Niño Southern Oscillation in the rainfall of Pucara basin and in the groundwater levels of the Punata alluvial fan in the Bolivian sub-Andes. Climate change and variability were assessed using the Quantile Perturbation Method, by detecting anomalous temporal changes in extreme quantiles of annual precipitation in the Pucara watershed and the correlation with groundwater levels in the Punata fan. The results show oscillatory behavior over periods of 28 to 33 years for the occurrence of wet and dry extremes at all studied meteorological stations. This suggests a similar oscillatory behavior of the groundwater levels; however, longer groundwater level observations are needed in order to confirm the link between precipitation variability and groundwater fluctuations. Local actors such as water managers, farmers and decision makers must take into consideration this climate variability in order to plan for these multi-decadal variations in response to the changes.

The skill, assumptions, and uncertainty of machine learning techniques (MLTs) for downscaling global climate model’s precipitation to the local level in Bolivia were assessed. For that, an ensemble of 20 global climate models (GCMs) from CMIP6, with random forest (RF) and support vector machine (SVM) techniques, was used on four zones (highlands, Andean slopes, Amazon lowlands, and Chaco lowlands). The downscaled series’ skill was evaluated in terms of relative errors. The uncertainty was analyzed through variance decomposition. In most cases, MLTs’ skill was adequate, with relative errors less than 50%. Moreover, RF tended to outperform SVM. Robust (weak) stationary (perfect prognosis) assumptions were found in the highlands and Andean slopes. The weakness was attributed to topographical complexity. The downscaling methods were shown to be the dominant source of uncertainties. This analysis allowed the derivation of robust future projections, showing higher annual rainfall, shorter dry spell duration, and more frequent but less intense high rainfall events in the highlands. Apart from the dry spell’s duration, a similar pattern was found for the Andean slopes. A decrease in annual rainfall was projected in the Amazon lowlands and an increase in the Chaco lowlands.

Land use and land cover change (LUCC) can affect the hydrological response time of rivers. However, it is difficult to generate robust and quantitative evidence of this impact at the catchment scale. This lack of evidence also affects the development of rainfall‐runoff models to make ex‐ante predictions. Here, we analyze high‐frequency observational data from a network of pairwise catchments in the tropical Andes and find a statistically significant impact of intensive land use on the hydrological response time, which can be used for regionalization. First, we isolated individual rainfall response events from 5‐min precipitation and discharge time series of 16 catchments (8 pairs). We then fitted unit hydrographs on these events to estimate the catchment response times. These response times were subsequently regionalized by, first, applying a forward stepwise regression to select statistically significant catchment characteristics including land use and land cover, then, fitting a linear mixed‐effects model with the selected characteristics to account for within‐site variability between pairs. We find that catchments with intensive land use have a significantly quicker response than their natural counterparts. Differences were often sub‐hourly, highlighting the value of high‐frequency monitoring. Forward stepwise regression identified only catchment area and intensive land use percentage (LUP) as statistically significant predictors. Model coefficients show that, even when considering other catchment characteristics, increasing intensive LUP decreases response times. This study provides solid evidence and a robust methodology to quantify the impacts of LUCC on catchment hydrology.

We assessed machine learning techniques (MLTs) for downscaling global climate model’s precipitation to local level in Bolivia. The skill, assumptions and uncertainty of the process are analyzed. Additionally, future projections are delivered. For that, an ensemble of 20 global climate models (GCMs) from CMIP6, with random forest (RF) and support vector machines (SVM), is used on four zones (highlands, Andean slopes, Amazon lowlands and Chaco lowlands). The downscaled series’ skill is evaluated in terms of relative errors. The uncertainty is analyzed through variance decomposition. MLTs’ skill is adequate in most cases, especially for the highlands and the Andean slopes. Moreover, RF tends to outperform SVM. Stationary assumptions are robust in the highlands and Andean slopes. Perfect prognosis assumption is poor in the highlands and Andean slopes; topographical complexity appears to be the reason. Regarding uncertainties, MLTs are the dominant source. In future projections, the highlands show shorter dry spell lengths with more frequent but less intense high rainfall events and higher annual rainfall. The Andean slopes exhibit an increased annual rainfall with a reduction in the high precipitation intensities but an increased frequency. The Amazon lowlands present a decrease in annual rainfall, and the Chaco lowlands an increase.