Table 3 - uploaded by Roberto Ranzi
Content may be subject to copyright.
Estimate of the correction factor Cs for each monitoring site and statistics of the resulting errors by considering the freshly-fallen snow water equivalent as the 'true' reference measurement.
Source publication
Precipitation measurements by rain gauges are usually affected by a systematic underestimation, which can be larger in case of snowfall. The wind, disturbing the trajectory of the falling water droplets or snowflakes above the rain gauge, is the major source of error, but when tipping-bucket recording gauges are used, the induced evaporation due to...
Contexts in source publication
Context 1
... correction factor Cs was calibrated for each monitoring site by minimizing the square bias of the individual measurements or the bias between the cumulative values ( Table 3). The Cs value appeared to always be higher in the latter case. ...
Context 2
... if V > 2.5, then Cs,w = 1.5 (Cancano P4) Since this second procedure is no more an at-site correction through a best fit value of Cs,w for each station, but has a more general climatological value, a slight loss of accuracy (bias) and precision (RMSE) was observed, as expected, although it has to be kept in mind that the dataset used for the climatological correction is smaller than the one used for the at-site correction, because wind data are available only for a subset of measurements. The results are still acceptable, though, as it can be seen by comparing the error standard deviation in Tables 3 and 5 and, for this reason, this second method is suggested in case both temperature and wind data are available. ...
Similar publications
Accurate characterization of precipitation P at subdaily temporal resolution is important for a wide range of hydrological applications, yet large-scale gridded observational datasets primarily contain daily total P. Unfortunately, a widely used deterministic approach that disaggregates P uniformly over the day grossly mischaracterizes the diurnal...
Citations
... The above findings highlight that gauge estimates are generally lower than plain reanalysis estimates in the mountains. It is important to take into account that automated measurements tend to underestimate the amount of precipitation, especially snow, depending on the type of gauges used and the wind speed (Grossi et al., 2017;Kochendorfer et al., 2022). Rasmussen et al. (2012) mentions errors from 20% to 50% for solid precipitation. ...
Seasonal snow is an important water source and contributor to river discharge in mountainous regions. Therefore the amount of snow and its distribution are necessary inputs for hydrological modeling. Recent research has shown the potential of the Sentinel‐1 radar satellite to map snow depth (SD) at sub‐kilometer resolution in mountainous regions. In this study we assimilate these new SD retrievals into the Noah‐Multiparameterization land surface model using an ensemble Kalman filter for the western European Alps. The land surface model was coupled to the Hydrological Modeling and Analysis Platform (HyMAP), a global flow routing scheme that provides simulations of routed river discharge. The performance with different precipitation forcing inputs, namely MERRA‐2 (with and without gauge based correction) and ERA5, was compared based on in situ precipitation and SD stations, with ERA5 leading to the best SD performance. The Sentinel‐1 based data assimilation (DA) results show small but systematic improvements for SD estimates, with the mean absolute error reducing from 36.4 cm for the open loop (OL) to 35.6 cm for the DA across all stations and timesteps, improving 318 out of 516 in situ sites. The DA updates in SD also result in enhanced snow water equivalent and discharge simulations. The median temporal correlation between discharge simulations and measurements increases from 0.73 to 0.78 for the DA. This study demonstrates the utility of the Sentinel‐1 SD retrievals to improve not only the representation of snow in mountain ranges, but also the snow melt contribution to river discharge, and hydrological modeling in general.
... These two signals are correlated with a Spearman rank correlation of = 0.58, ( = 0.0001), which is satisfying considering the lower elevation of the meteorological station and the disturbing effect of measurement errors. In a recent study by Grossi et al. (2017) biases between 15% and 66% were observed in solid precipitations measured at rain gauges belonging to the same region. On the other hand, summer ablation increased steadily with time (i.e., it became more negative) as shown in Fig. 5b, with smaller oscillations around the moving average with respect to . ...
... Thus, snow accumulates in the funnel, whose storage capacity is not usually large, and measurement is delayed until it thaws. Likewise, in hail conditions, much of the precipitation is lost due to bouncing of the hail in the funnel [89][90][91]. ...
Tipping bucket rain gauges (TBRs) continue to be one of the most widely used pieces of equipment for rainfall monitoring; they are frequently used for the calibration, validation, and downscaling of radar and remote sensing data, due to their major advantages—low cost, simplicity and low-energy consumption. Thus, many works have focused and continue to focus on their main disadvantage—measurement biases (mainly in wind and mechanical underestimations). However, despite arduous scientific effort, calibration methodologies are not frequently implemented by monitoring networks’ operators or data users, propagating bias in databases and in the different applications of such data, causing uncertainty in the modeling, management, and forecasting in hydrological research, mainly due to a lack of knowledge. Within this context, this work presents a review of the scientific advances in TBR measurement uncertainties, calibration, and error reduction strategies from a hydrological point of view, by describing different rainfall monitoring techniques, summarizing TBR measurement uncertainties, focusing on calibration and error reduction strategies, discussing the state of the art and providing future perspectives of the technology.
... Thus, 9 of 21 snow accumulates in the funnel, whose storage capacity is not usually large, and measurement is delayed until it thaws. Likewise, in hail conditions, much of the precipitation is lost due to bouncing of the hail in the funnel [90][91][92]. ...
Tipping bucket rain gauges (TBRs) have been, and apparently will continue to be one of the most widely used pieces of equipment for rainfall monitoring, being frequently used for the calibration, validation and downscaling of radar and remote sensing data, due to their major advantages–low cost, simplicity, and low energy consumption. Thus, many works have focused and continue to focus on their main disadvantage–measurement biases (mainly in wind and mechanical underestimations). However, despite arduous scientific effort, calibration methodologies are not frequently implemented by monitoring networks operators or data users, propagating bias in databases and in the different applications of such data, causing uncertainty in the modeling, management, and forecasting in hydrological research, mainly due to a lack of knowledge. Within this context, this work presents a review of the scientific advances in TBR measurement uncertainties, calibration, and error reduction strategies from a hydrological point of view, by describing different rainfall monitoring techniques in Section 2, summarizing TBR measurement uncertainties in Section 3, focusing on calibration, and error reduction strategies in Section 4, a discussion and perspectives in Section 5, and conclusions in Section 6, providing an overview of the of the state of the art and future perspectives of the technology.
... In particular, products such as IMERG, MERRA-2, or ERA sometimes show a certain underestimation of the snowpack, which makes these instruments unreliable for the purposes of a detailed analysis of snow depths [15,16]. Similarly, it is also complex for heated rain gauges to be able to count the snow water equivalent (SWE), since there are interactions with the wind that can generate underestimates in quantity when not properly shielded by special devices [17][18][19]. Precisely in relation to the possibility of snowfall being blown away by the wind, there are studies evaluating the preferential accumulation of wind-borne snow in mountain or glacial environments [20]. Because of these problems in accurately counting snow, recent efforts are being made to create models that can simulate snow cover and in some cases, predict it in relation to other climatic variables [21]. ...
... In this case, on the other hand, data were obtained from weather stations equipped with sonar to assess the depth of the snow and from manual weather stations, where the measurements were taken in the presence of the operator. Data from heated rain gauges were deliberately discarded because they are very often inaccurate due to the strong winds in a mountain environment, which generate significant underestimates when counting the snowpack thickness [17,27]. Moreover, in this part of the Apennines in central Italy, detailed snow cover studies had never been carried out, given the fragmentary and scarce data. ...
Research on solid precipitation and snow cover, especially in mountainous areas, suffers
from problems related to the lack of on-site observations and the low reliability of measurements,
which is often due to instruments that are not suitable for the environmental conditions. In this
context, the study area is the Monti Sibillini National Park, and it is no exception, as it is a mountainous
area located in central Italy, where the measurements are scarce and fragmented. The purpose of this
research is to provide a characterization of the snow cover with regard to maximum annual snow
depth, average snow depth during the snowy period, and days with snow cover on the ground in
the Monti Sibillini National Park area, by means of ground weather stations, and also analyzing
any trends over the last 30 years. For this research, in order to obtain reliable snow cover data,
only data from weather stations equipped with a sonar system and manual weather stations, where
the surveyor goes to the site each morning and checks the thickness of the snowpack and records,
it were collected. The data were collected from 1 November to 30 April each year for 30 years,
from 1991 to 2020; six weather stations were taken into account, while four more were added as
of 1 January 2010. The longer period was used to assess possible ongoing trends, which proved to
be very heterogeneous in the results, predominantly negative in the case of days with snow cover
on the ground, while trends were predominantly positive for maximum annual snow depth and
distributed between positive and negative for the average annual snow depth. The shorter period,
2010–2022, on the other hand, ensured the presence of a larger number of weather stations and was
used to assess the correlation and presence of clusters between the various weather stations and,
consequently, in the study area. Furthermore, in this way, an up-to-date nivometric classification
of the study area was obtained (in terms of days with snow on the ground, maximum height of
snowpack, and average height of snowpack), filling a gap where there had been no nivometric study
in the aforementioned area. The interpolations were processed using geostatistical techniques such as
co-kriging with altitude as an independent variable, allowing fairly precise spatialization, analyzing
the results of cross-validation. This analysis could be a useful tool for hydrological modeling of the
area, as well as having a clear use related to tourism and vegetation, which is extremely influenced
by the nivometric variables in its phenology. In addition, this analysis could also be considered a
starting point for the calibration of more recent satellite products dedicated to snow cover detection,
in order to further improve the compiled climate characterization.
... Snow measuring systems based on the quantification of precipitation must have heated systems, producing evaporation losses when precipitation is weak. In addition, strong winds decrease the snow collected in snow gauges when wind shields are not installed, causing losses of ~20%-50% (Grossi et al., 2017;Masuda et al., 2019;Rasmussen et al., 2012). Similar values have been found in the Cantabrian Mountains, where the underestimation of snow precipitation using the Spanish Meteorological Agency (AEMET) network is similar to that in the Pyrenees, but less than those in other areas of the Iberian Peninsula (Buisán et al., 2017(Buisán et al., , 2022. ...
... Another source of uncertainties is the measurement of precipitation itself since wind and evaporation can lead to an underestimation of precipitation (Rodda and Dixon, 2012;Grossi et al., 2017). However, the errors are generally higher for solid precipitation than for liquid precipitation and largest at windy sites (above 50 %) (Rodda and Dixon, 2012;Grossi et al., 2017). ...
... Another source of uncertainties is the measurement of precipitation itself since wind and evaporation can lead to an underestimation of precipitation (Rodda and Dixon, 2012;Grossi et al., 2017). However, the errors are generally higher for solid precipitation than for liquid precipitation and largest at windy sites (above 50 %) (Rodda and Dixon, 2012;Grossi et al., 2017). Since we use measurements from AWS2 in the valley, snowfall is less frequent. ...
The energy and mass balance of mountain glaciers translate into volume changes that play out as area changes over time. From this, together with former moraines during maximum advances, information on past climate conditions and the climatic drivers behind during glacier advances can be obtained. Here, we use the distributed COupled Snowpack and Ice surface energy and mass balance model in PYthon (COSIPY) to simulate the present state of an Italian glacier, named Fürkeleferner, for the mass balance years 2013–2017. Next, we investigate the local climate during the time of the last “Little Ice Age” (LIA) maximum glacier advance using COSIPY together with the LIA glacier outline retrieved from moraine mapping and a digital elevation model (DEM) adapted for the glacier’s geometry at the time of the LIA as a benchmark. Furthermore, the glacier’s sensitivity to future air temperature increase of +1 K and +2 K is investigated using the same model. For all simulations, meteorological data of closely located climate stations are used to force the model. We show the individual monthly contribution of individual energy and mass balance components. Refreezing during the summer months is an important component of the energy and mass balance, on average about 9 % relative to total annual ablation. The results from simulating past climate show a 2.8 times larger glacier area for Fürkeleferner during the LIA than today. This further implies a 2.5 K colder climate, assuming that the amount of precipitation was 10 %–20 % in excess of today’s value. Concerning further temperature increase of 2 K, the glacier would only consist of the ablation area implying sustained mass loss and eventual total mass loss. Even under current climatic conditions, the glacier area would have to decrease to 17 % of its current area to be in a steady state. We discuss the reliability of the results by comparing simulated present mass balance to measured mass balances of neighboring glaciers in the European Alps and with short-term measurements on Fürkeleferner itself. In conclusion, we are able to show how the glacier responds to past and future climate change and determine the climatic drivers behind.
... mountainous areas), which may cause erroneous performance assessments especially in cases of localized and severe events, and a too low fraction of heated rain-gauges, which may cause an underestimated and delayed detection of precipitation in case of snowfalls (e.g. Grossi et al., 2017;Mair et al., 2016). On the other hand, surface temperature T A B L E 1 RMSE and correlation index (R) computed for the three parameterizations of the soil temperature at 12.00 m depth against the measurements in Potsdam station, Germany sensors amount to about 1000 and include data from the high-resolution regional network Dewetra as well as from the SYNOP stations (that are not assimilated within SPHERA). ...
Abstract A new convection‐permitting regional reanalysis, SPHERA (High Resolution REAnalysis over Italy), has been developed over Italy and the surrounding seas using the COSMO model at 2.2 km horizontal resolution. The reanalysis system is nested in the global reanalysis ERA5; upper‐air and surface observations are assimilated at the convection‐permitting scale by the COSMO nudging scheme. Before the complete production of the hourly three‐dimensional fields and surface/soil parameters over the period 1995–2020, general issues regarding the reanalysis set‐up needed to be addressed over a shorter test period. These include the identification of the best approach to downscale the lateral boundary conditions from the global driver, and the definition of the bottom boundary condition related to deep soil temperature. With respect to the downscaling methodology, the results show a clear benefit in using lateral boundary conditions directly from the global ERA5, despite the large resolution difference between the two modes (1:15), instead of providing them from an intermediate resolution COSMO‐based reanalysis. Moreover, the soil bottom boundary condition for temperature is reconstructed from the shallower ERA5 soil, using a site‐dependent method based on a delayed running mean of the ERA5 temperature at the deepest soil level. Finally, an evaluation of SPHERA has been performed with respect to the skill in simulating daily precipitation over 2 years. Compared with ERA5, SPHERA shows a higher ability in simulating moderate and intense events, markedly during summer, in terms of skill scores, frequency of occurrence and bias.
... Because of slow falling, snow hydrometers are more susceptible to deflection by wind-induced turbulence around the gauge, making snowfall measurements prone to large systematic errors (Rasmussen et al., 2012). In windy conditions, the underestimation of snowfall accumulation frequently ranges from 20 % to 50 % or even higher and additionally depends on other variables, such as exposure and the type of rain gauge (Rasmussen et al., 2012;Buisán et al., 2017;Grossi et al., 2017). Other systematic error sources are related to physical processes, such as evaporation from a bucket, wetting, and splashing. ...
The RainGaugeQC scheme described in this paper is intended for real-time quality control of telemetric rain gauge data. It consists of several checks: detection of exceedance of the natural limit and climate-based threshold as well as checking of the conformity of rain gauge and radar observations, the consistency of time series from heated and unheated sensors, and the spatial consistency of adjacent gauges. The proposed approach is focused on assessing the reliability of individual rain gauge observations. A quantitative indicator of reliability, called the quality index (QI), describes the quality of each measurement as a number in the range from 0.0 (completely unreliable measurement) to 1.0 (perfect measurement). The QI of a measurement which fails any check is lowered, and only a measurement very likely to be erroneous is replaced with a “no data” value. The performance of this scheme has been evaluated by analysing the spatial distribution of the precipitation field and comparing it with precipitation observations and estimates provided by other techniques. The effectiveness of the RainGaugeQC scheme was also analysed in terms of the statistics of QI reduction. The quality information provided is very useful in further applications of rain gauge data. The scheme is used operationally by the Polish national meteorological and hydrological service (Institute of Meteorology and Water Management – National Research Institute).
... On the other hand, (Berndt andHaberlandt, 2018, Ly et al., 2011) argue that, topographical impact was indispensable for only temperature reconstruction at all temporal resolutions and station densities, but it was less clear for daily to monthly P. Furthermore, all spatial interpolation techniques can perform poorly in regions with insufficient high-elevation data, due to inaccurate estimation of local lapse rates (Ruelland, 2020). Considering all above reviews, it can be concluded that the basin average observed P estimates are underestimated owing to the under-catch P and unevenly distributed lowdensity (approx. 1 gauge/2235 km 2 ) climate station network especially at high altitude areas which are the main active parts of the basins (Azmat, 2015, Grossi et al., 2017, Ranzi et al., 2003, Winiger et al., 2005. Moreover, the P increases with elevation up to 4500 m in Himalaya mountainous range and then start to decrease (Immerzeel et al., 2015) whereas the climatic stations are installed up to 3000 m elevation. ...
The study presents a comprehensive assessment of the key terrestrial water budget (TWB) components and their recent temporal trends using observed and gridded datasets across Upper Jhelum Basin (UJB) for the period of 2003-2014. Different gridded datasets were used for precipitation-P (APHRODITE, CHIRPS, PGMFD, ERA-5, and MSWEP), evapotranspiration-ET (MOD16A2, SSEBop, GLDAS-NOAH, TERRACLIMATE and DOLCE), Gravity Recovery and Climate Experiment (GRACE) derived terrestrial water storage (TWS) and observed discharge (Q) measurements. Firstly, gridded P and ET datasets were assessed through detailed spatiotemporal analysis against reference datasets and considering their adequacy to close the TWB. The imbalance in TWB and associated uncertainties of each component were also computed to understand the potential of gridded datasets for water budget studies. Finally, TWB components were subjected to trend analysis to evaluate temporal trends and their statistical significance over the study period. Overall, ERA5-Pand DOLCE-ET relatively performed better and hence considered to be the most appropriate to present the hydrological spatiotemporal variability across UJB. The imbalance in water budget was mainly associated with higher uncertainties in P especially during wet periods. The results of temporal trend analysis showed a) insignificant mixed trends for annual P with increase and decrease in monsoon and winter P, respectively b) mixed yet statistically significant trends for ET in different parts of basin at annual and seasonal scale c) significant decreasing trends for TWS throughout the UJB in winter and annual scale d) increasing trends for Q in all months of the year except January and February. Despite the imbalance in water budget estimation due to large uncertainties in gridded P, ET, and TWS datasets, the study underlines the important information about TWB dynamics and their spatiotemporal variability over UJB which can be used as the knowledge base for further impact assessment studies.
