Precipitation fields in an alpine Mediterranean catchment: Inversion of precipitation gradient with elevation or undercatch of snowfall?

Abstract

The mean precipitation measurements in a Mediterranean Alpine catchment in Sierra Nevada show an inversion of the gradient with the altitude beyond a certain threshold. Is it due to a real pattern or it can be explained by systematic error of solid precipitation measurement in gauges? Can we assess climatic fields in an alpine catchment from gauge measurement? This article describes a research developed to answer both questions in the Alto Genil Basin. As commonly happens in most of the basins, the spatio‐temporal information from climate gauges is limited; therefore to reduce uncertainty in estimates of climatic fields, some secondary information should be introduced. Since orographic conditions clearly influence precipitation, the relationship between this climatic variable and elevation is usually included as secondary information into the estimates. However, while there is a clear relationship between temperature and elevation, the relationship between precipitation and elevation is not so simple. In this article the analysis of the data performed allow us to demonstrate that there is a real inversion of the gradient within this Mediterranean Alpine area as other authors previously pointed in some tropical and subtropical zones. The intensity of this phenomenon and the altitude threshold from which it appears can be altered as a consequence of the undercath of the solid precipitation. To estimate precipitation fields, we have used different hypotheses about the intensity of the undercatch taking into account empirical corrections obtained for nearby mountain ranges. An analysis of the sensitivity of the results to the assumed undercatch hypothesis shows that it is not possible to estimate properly precipitation fields (the sensitivity of the results to the adopted hypothesis is high) in these alpine areas if we only have information about the precipitation measurements at the stations.

A Preliminary Assessment of the “Undercatching” and the Precipitation Pattern in an Alpine Basin

“…For the RK approach, the difference between these coefficients is less significant: 1.78 for snowfall and 1.40 for rainfall. The RK approach reduces the negative gradient of precipitation initially predicted by the RK for altitudes above 1000 meters, due to the phenomenon of gradient inversion (decrement of precipitation with altitude above a specific elevation threshold, described in Reference [30]). That research made a preliminary correction of the experimental data in the Alto Genil basin by applying experimental correction functions proposed by Buisan et al (2017) [14] for certain Spanish mountain ranges.…”

“…Consequently, the monthly model was selected as the best approximation, and this was used to obtain the inputs of the hydrological model. For precipitation, the best approximation of the data was obtained with linear KED, although RK better reproduced the shape of the data (inversion of precipitation gradient) for the Alto Genil basin [30]. Figure 7 shows the estimated precipitation and temperature fields versus elevation for the Canales basin.…”

“…The ideal situation would be to have enough gauge measurements to make a distributed precipitation field without the necessity of making any estimation. The problem is that, in alpine regions, due to accessibility issues, the availability of data is usually reduced [30]. In cases where limited funds are available, the number of gauges is very limited.…”

“…To perform this estimation in an optimal way when very limited primary data are available, some secondary information such as altitude is useful for improving the assessment [39,40], reducing its uncertainty. In a previous research work [30], we applied these techniques to a case study that integrated our pilot. The geostatistical assessment employed in this previous work included several techniques (linear and quadratic kriging with external drift (KED) and regression kriging (RK)), whereby altitude was incorporated in different ways to improve the assessment of the precipitation fields.…”

“…It allows a preliminary global assessment of undercatch and precipitation patterns (distribution between solid and liquid phase and spatial gradient with the elevation) in alpine basins. We apply the study in the Canales basin (Sierra Nevada, Spain), where we analyze if the inversion of precipitation gradient described by Collados-Lara et al (2018) [30] is maintained with the obtained correction of the undercatch due to systematic errors. This inversion of gradient is defined as the change in the usual increment of precipitation with altitude to a decrement of precipitation with altitude above a specific elevation threshold.…”