The Pacific–Andean region in western South America suffers from rainfall data scarcity, as is the case for many regions in the South. An important research question is whether the latest satellite-based and numerical weather prediction (NWP) model outputs capture well the temporal and spatial patterns of rainfall over the region, and hence have the potential to compensate for the data scarcity. Based on an interpolated gauge-based rainfall data set, the performance of the Tropical Rainfall Measuring Mission (TRMM) 3B42 V7 and its predecessor V6, and the North Western South America Retrospective Simulation (OA-NOSA30) are evaluated over 21 sub-catchments in the Pacific–Andean region of Ecuador and Peru (PAEP). In general, precipitation estimates from TRMM and OA-NOSA30 capture the seasonal features of precipitation in the study area. Quantitatively, only the southern sub-catchments of Ecuador and northern Peru (3.6–6° S) are relatively well estimated by both products. The accuracy is considerably less in the northern and central basins of Ecuador (0–3.6° S). It is shown that the probability of detection (POD) is better for light precipitation (POD decreases from 0.6 for rates less than 5 mm day⁻¹ to 0.2 for rates higher than 20 mm day⁻¹. Compared to its predecessor, 3B42 V7 shows modest region-wide improvements in reducing biases. The improvement is specific to the coastal and open ocean sub-catchments. In view of hydrological applications, the correlation of TRMM and OA-NOSA30 estimates with observations increases with time aggregation. The correlation is higher for the monthly time aggregation in comparison with the daily, weekly, and 15-day time scales. Furthermore, it is found that TRMM performs better than OA-NOSA30 in generating the spatial distribution of mean annual precipitation.

Actual evapotranspiration (ETa) explains the exchange of water and energy between soil, land surface, and atmosphere. Despite its importance, it remains difficult to measure directly. Grasslands represent a widespread ecosystem for which further assessment of the measurement and estimation of ETa is needed. Thus, the objective of this study was to compare measurements and estimations of ETa in a mountain grassland ecosystem made using different approaches. The study was conducted in the Zhurucay Ecohydrological Observatory, located in the high Andes of Ecuador between 3,500 and 3,900 m a.s.l. The study area is a representative site of the páramo ecosystem, in which the vegetation mainly consists of tussock grasslands. ETa was measured or estimated using the following methods: eddy-covariance (EC), volumetric lysimeters (Lys), water balance (WB), energy balance (EB), the calibrated Penman-Monteith equation (PMCal), and two hydrological models [the Probability Distribution Model (PDM) and the Hydrologiska Byråns Vattenbalansavdelning model (HBV-light)]. During 1 year, precipitation (P) accumulated to 1,094 mm while ETa (measured with EC) accumulated to 622 mm (with ETa/P = 0.57). On a daily basis, the EC method measured average ETa rates of 1.7 mm/day. The best daily estimates according to percentage bias (pbias), normalized root mean square error (nRMSE), Pearson's correlation coefficient (r) and the volumetric coefficient (ve) came from the HBV-light model, followed by the PMCal and the PDM (pbias: −2 to −20%, nRMSE: 12–15%, r: 0.7–0.9, and ve: 0.7–0.8). On the other hand, the WB, EB, and Lys estimates showed a poor performance (pbias: −10 to −19%, nRMSE: 25–93%, r: −0.4 to 0.5, and ve: −0.5 to 0.7). As the methods used in this study are of different types (hydrological, micrometeorological, and analytical), their suitability and applications are discussed in terms of their costs, temporal resolution, and accuracy. This study identifies low-cost and easy-to-implement alternatives to EC measurements, such as hydrological models and the calibrated Penman-Monteith equation. This study also allows us to provide an increment of progress on the accurate closure of the water balance in grasslands.

The downscaling of global climate models (GCMs) aims at incorporating finer scale information to their horizontal resolution in order to represent regional and local processes better. There are two main approaches to downscaling: statistical (based on data relationships between synoptic atmospheric variables and observations of local variables) and dynamical (based on the modelling of regional atmospheric processes and land–surface interactions). In this study, some predictive capabilities regarding the generation of station‐scale mean monthly temperature and rainfall of both a statistical artificial neural network (ANN‐based) and a dynamical weather research and forecasting (WRF‐based) downscaling approach are assessed. We have devised two versions of the statistical downscaling approach. One of them includes regional orographic variables as predictors to allow for spatial extrapolation; the other is purely local. Historical observational data, from the period 1990 to 1999, of two major watersheds in the Ecuadorian Southern Andes, the Jubones and Paute river basins, were used. Since, to a certain extent, the value added by downscaling techniques can be attributed to terrain information, it is worth noting that some characteristics of the selected catchments (as notorious altitude differences and the presence of qualitatively different precipitation regimes) provide a scientifically interesting location for exploring how finer scale effects are captured. For this reason, we concentrate on the ability of downscaling techniques to reproduce seasonality. A decade of evaluation proved that both approaches were able to qualitatively describe precipitation and temperature seasonal variations for different regimes at representative weather stations. Furthermore, the seasonality of precipitation represented by both downscaling approaches surpassed the seasonality representation of reanalysis data. However, shortcomings on the estimates were found. Specifically, dynamical downscaled precipitation estimates were prone to overestimation. Despite the fact that the considered downscaling approaches are different in nature, their ability to represent the high spatio‐temporal variability in this region highlights the importance of evaluating their strengths and limitations.