## Conceptual Flow Model and Hydrogeological Analysis for Several Scenarios of Future Climatic Changes

As seen in Fig. 26.3, three sectors, as well as the discharge areas, are identified inside the Escusa aquifer. In the area of Escusa, the flow diverges toward Castelo de Vide and toward the Sever River (Rio Sever). In the Porto Espada sector, the predominant flow is toward the Sever River. Therefore, the flow from the Porto Espada and Escusa sectors contributes to the major discharge area of the aquifer, and the Sever River. Near Castelo de Vide, discharge is towards the low permeability lithologies in contact with the carbonate aquifer in that area. The remaining area of contact of the aquifer with adjacent lithologies is with "impermeable" series and thus no more outflow areas are considered. According the variation of recharge values, the groundwater divides are displaced toward the area of the river or the area of Castelo de Vide.

The average annual recharge for aquifer long-term water balance is about 450 mm/year. Considering that hydraulic head values in the discharge areas are also well-known, the conceptual flow model [6] can be expressed in terms of a steady state flow problem, for which only one unknown variable exists and that is the hydraulic conductivity. This problem can be solved using a numerical flow model. The solution provides a homogeneous equivalent hydraulic conductivity value [7] allowing a steady state characterization of the flow domain at a regional scale. This accommodates the long-term mass balance of the aquifer and sectors expressed in the conceptual flow model presented in Fig. 26.3.

o ton :ooo mo jooo moo oooo 7000 ¡sooo 9000 10000 11000 12000

Fig. 26.3 Schematic cross-section showing the predominant flow directions and discharge areas of the Escusa (Castelo de Vide) carbonate aquifer. The distances along the axis are given in meters. The biggest arrows represent the predominant flow directions. The small arrows, crossing the aquifer boundaries represent the position of discharge areas o ton :ooo mo jooo moo oooo 7000 ¡sooo 9000 10000 11000 12000

Fig. 26.3 Schematic cross-section showing the predominant flow directions and discharge areas of the Escusa (Castelo de Vide) carbonate aquifer. The distances along the axis are given in meters. The biggest arrows represent the predominant flow directions. The small arrows, crossing the aquifer boundaries represent the position of discharge areas

The hydraulic conductivity values calculated by the numerical and analytical solutions were compatible with a steady state description of the aquifer at the regional scale in terms of the existence of the defined sectors and in terms of the long-term water budget. The presented formulation to calculate hydraulic conductivity circumvents the need of knowing values that quantify the outflow volumes in the discharge areas of the aquifer.

The hydraulic conductivity values were calculated by means of two distinct theoretical conceptions for the interpretation of the aquifer hydraulic behaviour. In both cases, the known variables were the flow domain geometry and average hydraulic head values in discharge areas. The first solution was based on a numerical flow model where hydraulic conductivity was independent of hydraulic head and thus treating the aquifer as confined. The second solution which is analytical considers the aquifer whose water table is bounded by a free surface having a shape defined by the equilibrium between infiltration and hydraulic parameters characterizing each of the aquifer sectors.

The calculated values must be regarded as an equivalent hydraulic conductivity characterizing the entire flow domain [7]. A complete equivalence between the real heterogeneous medium and the idealized one is impossible. Therefore, the relation between the calculated equivalent hydraulic conductivity and the real values is defined, in a limited sense, according to certain criteria that must be equal for both media [18]. In the present case, the used criteria was based on flow equivalence and, additionally, on the definition of the global flow pattern of the aquifer.

The calculated values of hydraulic head using both the methods cannot be used in any other context other than the characterization of the aquifer steady state flow pattern at a regional scale. The description of the aquifer behaviour under specific stress conditions that are different from the average recharge values is impossible without a characterization of a parameter distribution considering the flow domain heterogeneity.

However, the obtained solutions allow the analysis of some crucial basic questions that shall be answered before the decisions required to build a more sophisticated model. This allows the analysis of more complex problems related to the parameters distribution in a flow domain where transient and diffuse flow are overlapped in a very complex pattern. First of all, it is possible to confirm the possible existence of the aquifer sectors proposed for defining the conceptual flow model. These sectors are present in an "artificial flow domain" similar to the real aquifer in terms of geometry, location of discharge areas and average water balance. Moreover, the global flow pattern can be described by different solutions based on a confined or unconfined description of the system.

Point values of hydraulic conductivity were calculated by the interpretation of pumping tests [8], which had permitted the estimation of hydraulic conductivity values characterising the fractured carbonate rock matrix and the nonfractured rock matrix. For the dissolution channels present in the carbonated rocks it is only possible to determine orders of magnitude for hydraulic conductivity in a simplified theoretical framework.

Groundwater undergoes geochemical evolution as it moves throughout flow systems. Therefore, hydrochemical and isotopic trends [1, 3] in each of the three sectors were also identified [11, 12, 20]. The results show that the predominant hydrochemical processes affecting water composition in the carbonate aquifer are the dissolution of carbonate rock minerals, mainly dolomite and accessory calcite. The saturation index (SI) values show that the Castelo de Vide sector tends to show a less accentuated trend to undersaturation with respect to calcite and dolomite than waters collected in the other two aquifer sectors (Fig. 26.4). At the same time, the TDS of samples taken in the Castelo de Vide sector are characterised by the highest values in the aquifer. This is reflected by the electrical conductivity (EC) values of water in this sector, whose average values are about 100 mS/cm higher than the values registered in the Escusa and Porto Espada sectors.

The described trends in the spatial distribution of EC and SI related with calcite and dolomite reflect the regional flow pattern defined in the conceptual flow model. Also, the observed hydraulic behaviour of the aquifer in the identified hydrochemical trend seems to be related to the time residence of water which must be longer in the Castelo de Vide sector due to the fact that the secondary outflow controlling the flow pattern in the NW area is toward relatively low permeable lithologies, that have a limited capacity to assimilate the transference's from the carbonate rocks. On the other hand, the flow toward the Sever River is more effective and thus the residence time of water flowing from the Escusa and Porto Espada sectors must be shorter, as shown by the presence of less mineralised waters and lower values of the SI with respect to calcite and dolomite.

In the Castelo de Vide sector, the residence time of water is longer due the low permeability of the lithologies receiving outflow from carbonate rocks in the secondary discharge area of the aquifer near Castelo de Vide. Therefore, the amount of total dissolved solids in water is more important than in the other aquifer sectors because the chemical processes of carbonate dissolution are closest to equilibrium. This is reflected by the highest values of EC in the Castelo de Vide sector. Due to the rapid outflows toward Sever River (Rio Sever), residence times are shorter in the Escusa and Porto Espada sectors. Here, the water is more undersatu-rated with respect to the dissolution of carbonate minerals than in Castelo de Vide sector, and the EC is also lower. The elevation of the recharge area in the Porto

Espada sector is about 100 m higher than in the other sectors. Depletion in 18O is observed allowing distinguishing the isotopic composition of water in this sector.

Another hydrochemical trend identified in the aquifer is related to 818O ratio values. Lower values of 818O represent a depletion of the 18O which is the heaviest isotope in relation with the lighter isotope 16O. This property is of particular utility in diverse hydrologic applications, namely in the identification of groundwater origin in aquifers characterised by the existence of recharge areas with different altitudes. That altitude effect was detected in the Castelo de Vide Aquifer, where measured values of 818O show that in the Porto Espada sector water is depleted about 0.3%c in 18O with respect to the Castelo de Vide and Escusa sectors [12]. Those changes in values of 818O are related to the fact that average altitude in the Porto Espada sector is around 650 m, and about 520-550 m in Castelo de Vide and Escusa sectors.

The identified trends of hydrochemical processes at regional scale which allow an indirect confirmation of the defined conceptual flow model for the aquifer are summarised in Fig. 26.4. Based on the conceptual model in Fig. 26.3 and the recharge-infiltration balance which varies from year to year, a simulation for the next 50 and 100 years was performed based on four climatic scenarios defined by the Hadley Centre for Climate Prediction and Research. Table 26.1 reflects the expected modifications in the precipitation values according the four scenarios, two of them for 50 years and the other two for 100 years. The values can be compared with the averages of a 40 years series (1959-1998), as shown in the same table. Figure 26.5 presents the average transference volumes from the aquifer to Sever River (upper diagram) and for the granitic rocks in contact with the aquifer in the area of Castelo de Vide (down diagram). In this figure, it can be noticed that the

Table 26.1 Monthly precipitation considering four scenarios defined by the Hadley Centre for Climate Prediction and Research for the latitude of the Escusa aquifer, two for the year 2050, two for the year 2100

Previewed monthly average (mm)

Table 26.1 Monthly precipitation considering four scenarios defined by the Hadley Centre for Climate Prediction and Research for the latitude of the Escusa aquifer, two for the year 2050, two for the year 2100

Previewed monthly average (mm)

Observed values (mm) |
2050 |
2100 | |||

Average of 40 years |
HadCM3 B2a |
HadCM3 A2c |
HadCM3 B2a |
HadRM2 | |

Jan |
117.02 |
117.21 |
106.84 |
81.43 |
44.38 |

Feb |
105.40 |
117.10 |
81.48 |
108.20 |
58.27 |

Mar |
72.83 |
91.47 |
55.13 |
98.41 |
94.72 |

Apr |
72.43 |
75.69 |
70.91 |
84.98 |
101.91 |

May |
68.59 |
83.34 |
65.44 |
76.82 |
102.88 |

Jun |
34.13 |
41.16 |
32.56 |
38.23 |
32.44 |

July |
7.19 |
6.16 |
5.46 |
6.27 |
6.15 |

Aug |
8.25 |
6.06 |
5.30 |
6.05 |
4.45 |

Sept |
45.13 |
30.69 |
29.96 |
30.05 |
13.74 |

Oct |
93.53 |
66.31 |
61.26 |
54.90 |
32.28 |

Nov |
114.39 |
88.19 |
73.44 |
54.90 |
35.57 |

Dec |
120.21 |
105.78 |
112.03 |
67.32 |
46.67 |

Total |
859.08 |
829.16 |
699.80 |
707.57 |
0.02 Fig. 26.5 Average transference values from the aquifer to the Sever River (upper diagram) and for the granitic rocks in contact with the aquifer in the area of Castelo de Vide (down diagram). Simulation based in the monthly average recharge volumes of 40 years (period 1959-1998) discharge to the Sever River, when compared with the discharge to the granitic rocks is significant. Even so, the more abundant vegetation in this area is surely related with this water transfer. In order to evaluate the impact of the future climatic scenarios, a simplification was made, considering that the recharge episodes in each month happen in a period of a quarter of the month, about a week. The representation of the discharge in the two areas (river and granitic rocks) was then compared with the four different future scenarios: (1) HadCM3 B2a (50 years), (2) HadCM3 A2c (50 years), (3) HadCM3 B2a (100 years) and (4) HadRM2 (100 years). The results are represented in Fig. 26.6 and the differences found between the actual discharge regimen and the simulated scenarios are more important than it seems by the observation of the figure. Respecting scenario 1 (HadCM3 B2a), for 50 years (see Table 26.1), there is a slight trend to an intensification of recharge in the first semester (January-July), which can represent an increment in the reserves in the dry period (June-September), with more capacity for abstractions, and a slight reduction in the transferences to the Sever River between September and December. There are not significant transfers of waters toward the granitic rocks in the Castelo de Vide area in this scenario. Concerning scenario 3 (HadCM3 B2a), for 100 years, it presents some similarity with scenario 1, but the increment of recharge in the beginning of the year only begins in the second half of the first semester. In this case, the increment of storage in the aquifer in the driest period is almost no sensitive. On the other hand, the decrease in the outputs between September and December is more evident on the water transfers to Sever River, which can be significantly affected in the last part of the year. Also in the last part of the year, the transfers to the granitic rocks decrease slightly. Fig. 26.6 Average transference values from the aquifer to Sever River (upper diagram) and for the granitic rocks in contact with the aquifer in the area of Castelo de Vide (down diagram). Simulation based in the monthly average recharge volumes for 40 years (1959-1998) and in the four established hypothetic future scenarios Fig. 26.6 Average transference values from the aquifer to Sever River (upper diagram) and for the granitic rocks in contact with the aquifer in the area of Castelo de Vide (down diagram). Simulation based in the monthly average recharge volumes for 40 years (1959-1998) and in the four established hypothetic future scenarios Concerning scenario 2 (HadCM3 A2c), for 50 years, there is a general trend to a reduction in the recharge. The same is true as the previous ones, the tendency increases in the end of the year. In this case, the storage in the aquifer during the driest months tends to a reduction. Scenario 4 (HadRM2), for 100 years, is the one that shows more differences in relation with the actual average trends. Excepting the period between March and May, in which a slight increment of recharge is expected, in all the other parts of the year there are a strong reduction in the flow of the Sever River, changing clearly the transfers from the aquifer (by less than half). Also the transfers to the granitic rocks in the area of Castelo de Vide are negligible when compared with the actual ones. |

## Post a comment