ESTAT / Doc. WR/00/3
Original in EN
Point 3 of the agenda
Proposed methodology for EUROWATERNET
for the calculation of fresh water resources
Pilot study: Selection of precipitation, inflow and outflow networks in Spain
Teodoro Estrela - EEA ETC/Inland Waters
Meeting of the
Task Force on Water Statistics
Ad-hoc group on water resources
Meeting of 7 and 8 March 2000
Bech Building - Room B2/404
WR/00/3P. 1
Introduction
The basis of EUROWATERNET is the information derived from existing national and/or regional monitoring networks within each member country. Member countries are asked to select stations that will form the basic precipitation, inflow and outflow networks. These networks are expected to be able to provide a general overview of the fresh water resources in each territory (country or river basin) at an European level.
A description is given below of a pilot design study in Spain for the three networks. Its basic aim is to confirm the suitability of the methodology described, in which a check is made to ensure that the simplifications have not adversely affected the reliability of the data.
Precipitation network
General characteristics
Only the rain gauge stations that fulfil the following requirements are selected:
- They must be in service
- They must have at least 30 years of complete daily precipitation data
Geographical distribution
The required number of rain-gauge stations should be geographically spread across a member country with a density of one station every 5,000 Km2. This spatial distribution must be as even as possible. The Spanish mainland covers a surface area of 494,000 km2. As a result, it holds that the precipitation network must be made up of about 100 points.
Distribution of the stations by altitude
The percentage of rain gauge stations lying between two particular elevations should be approximately the same as for the surface areas between these two elevations in the land under consideration. If, for example, the contours (lines joining point of equal altitude) are as indicated below, it must be possible to form approximately the same curve for the elevations of the selected rain gauge stations.
Figure. Distribution of stations with altitude
Relationship between surface and altitude in Spain
In accordance with this distribution 100 stations were selected. They can be seen in the following figure.
Figure. Stations eventually selected
Obtaining the precipitation
The next figure shows the map obtained by interpolating the average precipitation values for the 100 points selected. The interpolation method used was the squared inverse distance, the value of the three nearest points being considered in the interpolation.
Figure. Precipitation map obtained from the selected stations (Series 1940-96)
Estimation of the representativeness of the results
To check the suitability of the methodology used, the results shown above were compared with those obtained for all the stations on the Spanish network (about 10.000) and which were calculated with a view to preparing the Spanish White Paper on Water.
Figure. Precipitation map obtained from the data for all the meteorological stations in Spain. Source: Spanish White Paper on Water (Series 1940-96)
It was revealed that the greatest differences are about 20% in the smaller basins. The next figure shows the mean annual precipitation values in Spain obtained considering all the meteorological stations (peninsula) together with those obtained from the selection chosen (sel100).
Figure. Comparison of the development of the average value for the mean annual precipitation in Spain, considering all the meteorological stations together with those obtained from the selection chosen.
Internal inflow network
General Characteristics
An initial selection of the stations was made according the following criteria:
- The river regime at the station must be as natural as possible, with little influence from human activities (e.g. abstraction or regulation)
- Stations is intended to have a minimum record period of 25 years from 1960 to 1990.
- Stations must be active
Geographical distribution
Considering a relationship of ¼ between number of gauging stations and rainfall stations, it can be deduced a control point density of one for every 20,000 km2 i.e. the natural internal inflow network must consist of about 25 points.
Complying with the non-altered regime requirement and trying to obtain as even a geographical distribution as possible, the 28 stations that are shown in the following figure were selected. The stations considered are located for interpolation in the centre of gravity of the catchment areas that drain into them, and that is how they are portrayed in the figure.
Figure. The gauging stations and basins that were eventually selected. Note: The position of each station is not the real one, but the centre of gravity of its drainage basin.
Obtaining the Natural Internal Inflow
The next figure shows the map that is a result of interpolating the average runoff values for the 28 points selected. The interpolation method used was the squared inverse distance , the three nearest points being considered when interpolating the value.
The runoff map obtained from the selected stations (Series 1940-96)
Estimation of the representativeness of the results
To check the suitability of the methodology used, the results shown above were compared with those calculated with a view to preparing the Spanish White Paper on Water.
Figure. Runoff map obtained by mathematical modelling.
Source: Spanish White Paper on Water (Series 1940-96)
There is a 6 % difference between the value for the whole of the Spanish mainland obtained from the selection made and the value obtained with the mathematical model.
The next figure shows the annual values for the mean runoff in Spain, as obtained from the mathematical models, together with the results given by the selection made, in accordance with the methodology described here.
Figure. Comparison of the development of the average value for the mean annual runoff in Spain, as obtained with from mathematical modelling, and the results from the natural regime stations that were selected.
Actual outflow network
Selection of stations
A selection of the stations that fulfil the following requirements, was made:
- The gauging stations must either be located at the points lying closest to the mouth or at those where the river is administered by a different authority (example: the frontier between Spain and Portugal).
- Stations is intended to have a minimum record period of 25 years from 1960 to 1990.
- Stations must be active
Fourteen stations were selected and these are shown in the following figure
Figure. Selected stations
Obtaining the Actual Outflow
In the following table, actual outflow is shown in the major river basins in Spain. For the whole area of the river basin, a runoff equal to the measured in the representative station has been adopted.
River Basin / Surface (km2) / Actual outflow (mm)[1] / Actual outflow (hm3)[2] / TypeNorte I / 17600 / 534 / 9398,40 / Sea
Norte II / 17330 / 876 / 15181,08 / Sea
Norte III / 5720 / 922 / 5273,84 / Sea
Duero / 78960 / 109 / 8606,64 / Portugal
Tajo / 55810 / 138 / 7701,78 / Portugal
Guadiana I / 53180 / 27 / 1435,86 / Portugal
Guadiana II / 7030 / 219 / 1539,57 / Sea
Guadalquivir / 63240 / 44 / 2782,56 / Sea
Sur / 17950 / 61 / 1094,95 /
Sea
Segura / 19120 / 2 / 38,84 / SeaJúcar / 42900 / 14 / 600,60 / Sea
Ebro / 85560 / 127 / 10866,12 / Sea
CI Cataluña / 16490 / 96 / 1583,04 / Sea
Galicia Costa / 13130 / 898 / 11790,74 / Sea
Peninsula / 494020 / 158 / 77894,02 / Sea+Portugal
Peninsula / 494020 / 122 / 60149,74 / Sea
Peninsula / 494020 / 36 / 17744,28 / Portugal
Table. Aggregate value of the actual outflow for the main basins (Series 1975-96)
Actual outflow to sea is, in consequence, about 60.000 and actual outflow to Portugal is about 18.000 hm3 /year
Estimation of the representativeness of the results
In the Spanish White Paper on Water an average annual yield into the sea of about 64.000 hm3/year is considered, whereas the average annual yield into Portugal (Duero+Guadiana I+Tajo) is about 21.000 hm3/year. In view of the fact that the aforementioned figures have yielded errors of less than 10 % in estimating the actual outflow, it is considered that the methodology used is suitable enough.
An annual consumption demand of 26.000 hm3/year would be obtained for Spain by calculating the difference between the internal inflow (211 mm that is equivalent to 104.238hm3/year) and the actual outflow (77.894 hm3/year).
WR/00/3P. 1
[1] Representative station value in each major basin
[2] Result of using representative station yield for the whole area of the basin