Modélisation sur Enxoé - Lisbonne, 9-11 mars 2011

Transcription

1 Modélisation sur Enxoé - Lisbonne, 9-11 mars 211 1

2 Enxoé watershed Description Area : 6km2 River Length: 9km Altitude Range: 16 35m Annual Precipitation: 5mm Pressures: 5% NoIrrigAgriculture/Pasture; 35% Olive Trees; High animal presence (cattle, sheeps) low urban presence water presence/management (for small reservoirs) Pasture/agric. Permanent pools Olive tree Water retained in reservoirs 2

3 Montado Land Uses December May October 3

4 Land Uses Irrigation Olive Small Farms 4

5 Data Data for model implementation DTM NASA 9m Land Use Corine 2 Soil Texture EU soil Precipitation 5

6 Data Data for validation Collected during the project Neighbour watersheds (flow) Reservoir volumes and discharges 6

7 Watershed 65 basins 161 hru Auto fertilization (maximum 2kgN/ha.year) 7

8 SWAT - Calibration Ground water behaviour Impact on daily hydrogram Default Calibrated Porous Media Shallow Aquifer Gw_delay River Alpha_BF Soil depth Impact on monthly flow 8

9 Comparison with long term historic flow data outside Enxoe Lack of historic in enxoé river basin 1 month data Aguaflash December 29 and from October 21. For checking long term trends and temporarility Run SWAT to neighbour watersheds and cal/validate to long term data Monte da Ponte with 455 km 2 ; 1-3m alt.; 45km river Limas with 97 Km 2 ; 1-25m alt.; 25km river Enxoé is 6 Km 2 ; 18-3m alt.; 9km river 9

10 Comparison with long term historic flow data outside Enxoe Monthly Flow Monte da Ponte Flow Flow (hm3/month) Jan-1959 Jan-196 Dez-196 Dez-1961 Dez-1962 Dez-1963 Dez-1964 Dez-1965 Dez-1966 Dez-1967 Dez-1968 Dez-1969 Dez-197 Dez-1971 Dez-1972 Dez-1973 Dez-1974 Dez-1975 Dez-1976 Dez-1977 Dez-1978 Dez-1979 Dez-198 Dez-1981 Dez-1982 Dez-1983 Dez-1984 Dez-1985 Dez-1986 Dez-1987 Dez-1988 Dez-1989 Dez-199 Dez-1991 Dez-1992 Dez-1993 Dez-1994 Dez-1995 Dez-1996 Dez-1997 SWAT Data R2:.8 Nash:.68 Data SWAT Date R2:.77 Nash:.74 Flow (hm3/month) Jan-1982 Jan-1983 Jan-1984 Dez-1984 Dez-1985 Dez-1986 Dez-1987 Limas Flow Dez-1988 Dez-1989 Dez-199 Dez-1991 Dez-1992 Dez-1993 Dez SWAT Data Month

11 Comparison with long term historic flow data outside Enxoe Flow Frequency Monthly flow (m3/s) Rio Cobres - Monte da Ponte Station MPonte Data.1 % 1% 2% 3% 4% 5% 6% 7% 8% 9% 1% exceedance frequency Monthly flow (m3/s) Rio Cobres - Monte da Ponte Station % 1% 2% 3% 4% 5% 6% 7% 8% 9% 1% exceedance frequency MPonte SWAT 11

12 Comparison with indirect data inside Enxoe Enxoe Reservoir balance 13 years data Discharges Water abstraction Volume change Meteorology Evaporation Precipitation Compare to SWAT model Inflow and Mohid Model inflow 12

13 Comparison with indirect data inside Enxoe Reservoir Input 2.5 hm3/month From Reservoir Balance SWAT -.5 Set-25 Dez-25 Mar-26 Jul-26 Out-26 Jan-27 Abr-27 Ago-27 Nov-27 Fev-28 month Jun-28 Set-28 Dez-28 Mar-29 Jul-29 Out-29 Jan-21 13

14 Water Quality Simulation - SWAT Nitrate (mgn/l) SWAT Data Data Low concentrations when compared to intensive agriculture areas 14

15 Water Quality Simulation - SWAT SST (mg/l) SWAT Data Data 15

16 Validation in long term hydrodinamics and qualitatively water quality Balances 16

17 Water Balance mm/year Hydrol Year 2% goes in river Precipitation Flow 17

18 Precipitation 5 mm Water Balance % 9% 8% % 7% 6% 5% 4% 3% 2% 1% % % 8% Hydrol Year Balance Runoff Percolation Lateral Flow EVPTP Runoff 7% Percolation 6% 5% Evapotranspiration 4% 3% 2% 1% 18 %

19 Nitrate Balance NO3 pathaways in agriculture land kg/ha.year >2 kg/ha.year ->1 kg/ha.year 6% 5% Hydrol Year Balance NO3 Runoff NO3 GW Runoff Lateral Flow Groundwater 4% NO3 Lateral Flow 3% 2% 1% % 19

20 Montado December May October Irrigation Olive Small Farms 2

21 Nitrates vs. Water yield et Précipitations y =.12x R 2 = Nitrate ton/year Precipitation that does not get to river and not transport nitrate Precipitation mm/year Nitrate ton/year y =.22x R 2 = River Flow mm/year 21

22 Soil Loss Soil Loss tonss/ha.year >2 ton/ha.year ->5 ton/ha.year 1% 9% 8% 7% 6% 5% Hydrol Year % of annual mass occuring in 2 to 4 days in year % of annual mass 4% 3% 2% 1% %

23 MES vs. Water yield et Précipitations y = x R 2 = SST ton/year y = 5.255x R 2 =.9539 Precipitation mm/year SST ton/year River Flow mm/year 23

24 mm/year tonss/year tonn/year Hydrol Year Hydrol Year Hydrol Year Precipitation Flow Suspended Sediment Load Nitrate Load 24

25 Flow Enxoé 25 2 Dry year, Wet year, _Hum 26_Average 1991_Dry 29 Flow (m3/s) Flow (m3/s) Suspended Sediment (m3/s) Ago 22-Ago 21-Set 21-Set 21-Out 21-Out 2-Nov 2-Nov 2-Dez 2-Dez 19-Jan 19-Jan 18-Fev 18-Fev 2-Mar Nitrate Enxoé 2-Mar Date 19-Abr 19-Abr 19-Mai 19-Mai Suspended Sediment Enxoé Date 18-Jun 18-Jun 18-Jul 18-Jul 17-Ago 17-Ago 16-Set 16-Set 16-Out 1995_Hum 26_Average 1991_Dry _Hum 26_Average 1991_Dry Out 15-Nov 15-Nov 22-Ago 21-Set 21-Out 2-Nov 2-Dez 19-Jan 18-Fev 2-Mar 19-Abr 19-Mai 18-Jun 18-Jul 17-Ago 16-Set 16-Out 15-Nov 25

26 Flow decomposition 8 7 Daily discharge (m3.s-1) Runoff Lateral flow Groundwater 1 Oct-9 Dec-9 Feb-1 Apr-1 Jun-1 Aug-1 Oct-1 Time (days) 26

27 Mohid Land 6x4 cells 2mx2m 12 layers soil.5-3m Developments Vegetation Property transport Erosion/deposition Pesticide application/itenerary Data for implementation Data for validation 27

28 Comparison Reservoir Input hm3/month MohidLand SWAT Reservoir Balance -.5 Nov-27 Fev-28 Jun-28 Set-28 Dez-28 Mar-29 Jul-29 Out-29 Jan-21 month Jan-Fev 29 SWAT ET 15-2% Precip MOHID LAND ET 3%-5% Precip 28

29 SST first results Mohid Land Data Data 29

30 December 29 events 3

31 Water Depth December 29 events Probe MOHID 31

32 December 29 events Where did the water went to? Peaks inside soil surface layer MOHID Rain peaks in soil not producing saturation Rain peaks producing saturation 32

33 Water Depth December 29 events Reality infiltrated less and runoff to river Soil Saturation Model infiltrated all and no runoff to river Probe MOHID 33

34 December 29 events SWAT MOHID 34

35 Runoff Processes Permeable soils Precipitation Rate Less permeable soils Precipitation Rate Conductivity Rate Conductivity Rate Processes occur also for same soil depending on precipitation rates 35

36 Sensitivity analysis Mohid Land a physic tool to understand dynamics Sensibility Tests Lowering conductivity (promoting infiltration excess) Reducing soil depth (promoting saturation excess) Reducing surface rugosity (promoting flow) Bibliography Impermeabilization, soil surface sealing? 36

37 Soil sealing Processes Then dries out and forms a crust with low permeability Assouline, S. and Y. Mualem, 2, Modeling the dynamics of soil seal formation. Analysis of the effects of soil and rainfall properties. Water Resour. Res. 6: Assouline and Y. Mualem, 21, Soil seal formation and its effect on infiltration. Heterogeneous versus homogeneous seal approximation. Water Resour. Res. 37(2): Assouline, S. and Y. Mualem, 22, Infiltration during soil sealing: The effect of areal heterogeneity of soil hydraulic properties. Water Resour. Res. 38(12): Assouline, S. and Y. Mualem, 23, Effect of rainfall induced soil seal on the soil water regime: Drying interval and subsequent wetting. Transport in Porous Media 53: Assouline, S. and Y. Mualem, 26, Runoff from heterogeneous small bare catchments during soil surface sealing. Water Resour. Res. 42, W1245, doi:1.129/25wr

38 Impermeability Results December 29 events Reality Some infiltration occurs Reality Behaving impermeable Probe All watershed saturated Mohid orig Mohid imp Hint: Watershed is not impermeable but behaving has impermeable 38

39 Enxoé behaviour in December peaks Related to impermeability or resistence to infiltration (as impermeable) It could be ponctual behaviour related to high precipitaion event in October (65mm in 2h) Next actions See if this behavious appear in other level measures records. Ponctual or dominant behaviour Models are giving same long SWAT term trends MOHID of the system and with Mohid Land we may study event dynamics 39

40 What data says Salinity Enxoé Autom Sampling Profond. Salinity October 21 Profond (m) h 1h 1h 12h 3d Salinity (ppt) : : : : : : : : : Data : : : m : : : : Fev : Salinity Enxoé 6h 4h 1d 1d 6d Profond. Autom Sampling Salinity ppt : : : : : : : : : : : : : : : : : 4 Data

41 What data says Salinity Enxoé December 29 Profond. Salinité h 2h prof (m).3 ppt.2 2d.1 Data 41

42 Conclusions Concept model Long term functioning Precipitation 5mm Runoff 1% Percolation 1% 1 kgn/ha Evapotranspiration 8% 2% Nitrate loss is low consistent with extensive agriculture and data in river Soil loss is not high consistent with gentle slopes, natural vegetation and data in river. And arrives in few events. But high percentage of anual load arrives in 2-4 days..4 ton/ha 1-2 kgn/ha.6 kgn/ha 8% 1% 1%.4 kgn/ha Floods Floods rise in 1h 6h and fall in 1h-1d In flood arrives water with lower salinity (mainly runoff) but couple of days recovers initial salinity (groundwater replaces water). Soil sealing may be important for decreasing permeability. Models SWAT good results in long term dynamics and qualitativeli in WQ Did not managed to show Dec9 peaks (CN, hydrologic group, lower drained soil..). SWAT will not solve Enxoé peaks (hours). Mohid Land good results in long term dyanmics Developing tool to understand watershed dynamics physically and in events 42

43 Next Actions Continue Sampling Automatic Sampler protection against floods and mices (pictures) Continue modelling Pesticide in SWAT (but it is not a issue) Sediment and pesticide in Mohid Land Impermeability continues? 43

44 Localization 44

45 45