Integrating wetlands and riparian zones in regional hydrological modelling Fred Hattermann, Valentina Krysanova & Joachim Post Potsdam Institute for Climate Impact Research
Outline Introduction Model concept: - the SWIM model - the riparian zone module Case study area: - the Nuthe basin Effects on: Conclusions/outlook - water balance - nitrogen balance
Motivation 72 soil types Alluvium Periglacial Loess Mountains Soil types Processes in wetlands and groundwater are often not taken into account in integrated water resources management (or not understood) Wetland processes are underrepresented in most hydrological models Wetlands may have a large impact on water quality and quantity They have to be taken into account in new national and European programmes, like the Water Framework Directive
Water fluxes at the catchment scale Riparian zones and wetlands serve as an interface between upland and river network, 1) they interacts with groundwater, 2) lateral fluxes from upland pass through riparian zone
Soil and Water Integrated Model (SWIM, Krysanova et al. 1998) Climate: solar radiation, temperature, & precipitation Elbe River Basin Hydrological cycle Soil profile Shallow groundwater Deep groundwater A B C Crop/vegetation growth LAI Biomass Roots Nitrogen cycle N-NO 3 N o-ac N o-st N res Phosphorus cycle P lab P m-ac P m-st Land use pattern & land management for application in river basins and at the regional scale P org P res Brandenburg
Hydrological cycle in SWIM: vertical fluxes net radiation air temperature relative air humidity wind speed surface roughness LAI field capacity passage time per layer hydraulic conductivity saturated conductivity percolation Evaporation Transpiration soil water content? Precipitation Groundwater retention coefficient Surface runoff Subsurface runoff capillary rise land use type hydrol. soil type management slope drainable water from the saturated zone drainage porosity slope length Groundwater flow? Soil and groundwater interact only via percolation and capillary rise
Lateral fluxes in SWIM basin river routing (water, N, P, sed.) sub-basins hydrotopes? aggregation of water, N, P cycling, lateral flows vegetation growth? All lateral fluxes from a subbasin come directly to the river network
Model Extension Model definition: A riparian zone or wetland is defined as a hydrotope with shallow g-w table (where plant roots can reach groundwater) having lateral inflow from upland areas The main changes introduced in the model: A. implementation of daily groundwater table dynamics at the hydrotope level and soilgroundwater interaction, B. implementation of nutrient retention in groundwater and interflow (residence time and denitrification), C. implementation of water and nutrient uptake by plants from groundwater in riparian zones and wetlands.
A. Groundwater dynamics at the hydrotope level New approach in SWIM: water table is defined daily for each hydrotope automatic calibration (Hattermann et al. 2004) Before in SWIM: water table is defined daily for each subbasin 10T α = 2 µ L
B. Nitrogen Retention in Groundwater and Interflow ρ c u Dg σ + c c + λc + t R R R c = concentration n = eff. porosity m = aquifer thickness λ = turnover coeff R = faktor of retardation Simplification: Full mixture during the transport process. Residence time is normally distributed. Linear degradation. dct dt = C t, in Ct, out λ C t C t = KC t, out 1 (1/ K+ λ) t (1/ K+ λ) t Ct, out = Ct, in (1 e ) + Ct 1, oute 1+ Kλ q mn f R ( c c ) = 0 in Classical approach: the convection-dispersion equation But: a) it is nonlinear and has to be solved numerically b) high data demand => Basic equation K = mean residence time, λ = turnover coefficient, C = concentration
B. Nitrogen Transport and Retention Deposition
B. & C. Nitrogen retention and plant uptake: Parameter estimation L = K = n i= 1 dz i dz n i i= 1 vs ( zi ) k * J ( z) v s ( z) = S The distance L to the river is calculated following the gradient in groundwater table to the river. The mean residence time K is a function of flow path, permeability, porosity, and gradient in groundwater table for subsurface flow and gradient in topography and Manning s roughness for surface flow. It can be calculated using the seepage velocity v s (m d -1 ), where k in m d -1 is the hydraulic conductivity of the spatial unit z, J the dimensionless hydraulic gradient, and S the specific yield (average ~40 years, up to > 1000 years). x The turnover coefficient λ is a function of redox potential and carbon concentration of the catchment sediments. It was calibrated using data from Wendtland et al. (1993) for the Elbe basin as initial values (a half-life time of nitrate N between 1 and 3 years, which corresponds to λ values between 6 10-4 d -1 and 2 10-3 d -1 ). The maximum N uptake by plants from groundwater and interflow is limited by the available lateral discharge and was calculated using the flow accumulation method which calculates the amount of nutrients flowing through a spatial unit of the catchment following the groundwater gradient, and by the plant demand using a resistance function. y
Model concept Gauge station observation well precipitation station Nuthe climate station river system Berlin Nuthe basin Stepenitz The Nuthe Basin (1933 km2)
The Nuthe Basin: Groundwater Table Groundwater contour map < 40 m Gwstation Fgw_fein River Model concept Gwstation Fgw_fein Fgw Ges_elbe_poly < 400 401 520 521 640 641 760 761 880 881-1000 1001-1120 100 m >160 m [m] Main river FgwMain basins Ges_elbe_poly Observation wells 3000 3 0 3000 3 Kilometers Kilometers < 400 401 520 521 640 641 760 761 880 881-1000 1001 1120 1121 1240 1241 1360 1361 1480 1481-1600 > 1600
Spatial heterogeneity and delineation of hydrotopes Soil Landuse Groundwater table Geology Mean residence time / flow 23.419 distance hydrotopes etc.
River Discharche at Gauge Babelsberg (daily N&S 0.62) 35 30 25 Q simulated (daily) Q observed (daily) 20 15 10 5 0 4/1/88 1/1/81 4/1/81 7/1/81 10/1/81 1/1/82 4/1/82 7/1/82 10/1/82 1/1/83 4/1/83 7/1/83 10/1/83 1/1/84 4/1/84 7/1/84 10/1/84 1/1/85 4/1/85 7/1/85 10/1/85 1/1/86 4/1/86 7/1/86 10/1/86 1/1/87 4/1/87 7/1/87 10/1/87 1/1/88 900 800 700 Q simulated (monthly) Q observed (monthly) 600 500 400 300 200 100 0 1 7 13 19 25 31 37 43 49 55 61 67 73 79 85 91
Evapotranspiration and groundwater recharge River Groundwater recharge [mm/a] Evapotranspiration [mm/a]: Additional evapotranspiration from groundwater in wetlands is about 24 % of the total plant water uptake (~48 % of river discharge)
Groundwater table dynamics in a hydrotope water table [m] evapotranspration / recharge [mm] 4 3 2 1 0 150-1 8 200-2 -3 6 250 21.1.74 21.1.75 21.1.76 20.1.77 21.1.78 21.1.79 21.1.80 20.1.81 21.1.82 [m] [mm] 4 0 50 100 precipitation [mm] Precipitation, 2 evapotranspiration, groundwater 0 recharge and groundwater table -2 dynamics in a hydrotope. 0 evapotranspiration recharge water table simulated 2 water table observed precipitation 4-4 12 01.06.81 01.12.81 01.06.82 01.12.82 01.06.83 01.12.83 01.06.84 01.12.84 01.06.85 GW +rip GW -rip AET -rip AET +rip diff (AET) 6 8 10
Groundwater table dynamics in 9 observation wells 57.5 [meters above sea level] 55 52.5 50 47.5 45 42.5 40 37.5 35 32.5 30 27.5 Jun-1981 Jun-1983 Jun-1985 Jun-1987 Jun-1989 Jun-1991 Top: Groundwater dynamic simulated (black lines) and observed. Right: Location of the observation wells. 1 37421572 38431586 2 37431610 3 39461225 4 39441712 5 38431500 6 37441780 7 37441880 8 36441971 9 sim-changed2 sim-changed3 simulation sim-changed1 112 80.00 9 38 16 8 13 1 3 7 2 1 6 5 4 Gwstation Fgw_fein Fgw < 400 401 520 521 640 641 760 761 880 881-1000
Nitrate Leaching into Riparian Zones [kg/ha] Nitrate-N input from interflow to riparian zone Nitrate-N input from groundwater to riparian zone [kg/ha]
Nitrate Concentration in River 2.5 2.0 a nitrate sim nitrate obs [mg/l] [mg/l] [mg/l] 1.5 1.0 0.5 0.0 2.5 2.0 1.5 1.0 0.5 0.0 2.5 2.0 1.5 1.0 0.5 b c N surface N interflow N base flow min max 10 th 90 th median Nitrate N concentration: simulated and observed Nitrate-N from two major flow paths Evaluation of the uncertainty of the results (Latin Hypercube method, variation of 28 model parameters for hydrology and nutrient retention): 100%, 80% and median 0.0 1/1/93 7/1/93 1/1/94 7/1/94 1/1/95 7/1/95 1/1/96 7/1/96 1/1/97 7/1/97 1/1/98 7/1/98
Simulated Nitrate-N uptake in the basin and in riparian zones N plant uptake [kg/ha] Additional uptake [kg/ha]
The effect of additional N uptake in riparian zones on N concentrations in the river [mg/l] river nitrate concentration 2.5 2 with uptake in riparian zones without uptake in riparian zones 1.5 uptake by plants difference 1 0.5 0 1/1/93 7/1/93 1/1/94 7/1/94 1/1/95 7/1/95 1/1/96 7/1/96 1/1/97 7/1/97 1/1/98 0 0.4 0.8 1.2 1.6 2 difference [mg/l]
Summary and Conclusions A correct representation of spatial heterogeneity is necessary to reproduce the water and nutrient fluxes at the basin scale. Riparian zones and wetlands have a high potential to reduce river flow (additional evapotranspiration) and nutrient loads (additional plant uptake) to surface water. In our case study additional evapotranspiration of about 24 % and additional nitrate uptake of about 6 % were simulated (~ 48 % of river discharge and ~24 % of river load). Ecohydrological models integrating relevant hydrological, biogeochemical and vegetation processes at the river basin scale can serve as a basis to investigate land use and climate changes impacts at the regional scale.
Q difference [m3/s] Model concept Outlook 3500 3000 2500 2000 1500 1000 500 0 Qsim rip+ Qsim rip- difference 700 600 500 400 300 200 100 0 01.01.87 01.04.87 01.07.87 01.10.87 01.01.88 01.04.88 01.07.88 01.10.88 01.01.89 Q [m3/s] 01.04.89 01.07.89 01.10.89 01.01.90 01.04.90 01.07.90 01.10.90 01.01.91 01.04.91 01.07.91 01.10.91 01.01.92 01.04.92 01.07.92 01.10.92 4000 3500 3000 2500 2000 1500 1000 500 0 1/1/81 7/1/81 1/1/82 7/1/82 1/1/83 7/1/83 1/1/84 7/1/84 1/1/85 7/1/85 1/1/86 7/1/86 1/1/87 7/1/87 1/1/88 7/1/88 1/1/89 7/1/89 1/1/90 7/1/90 1/1/91 7/1/91 1/1/92 7/1/92 beobachtet observed simuliert simulated river discharge [m3/s]