Planning with water

Planning with water - an overview : Planning with water - an overview Paul van Walsum


Overview : Overview introduction regional influencing through GW andamp; SW methods for decision support influence matrix method embedding method


Regional influencing through GW & SW : Regional influencing through GW andamp; SW pressure wave droplet movement


Regional influencing, matrix : Regional influencing, matrix


Regional influencing, cross section : Regional influencing, cross section


SIMGRO for the regional hydrology : SIMGRO for the regional hydrology


Methods for decision support : Methods for decision support simulation models optimization models linked optimization-simulation models


Planning with water, ‘conventional style’ : Planning with water, ‘conventional style’ simulation effects on objectives Stakeholders suggest measures communication


Planning with water, ‘inverse approach’ : Planning with water, ‘inverse approach’ optimization Stakeholders: targets on objectives options for measures measures communication


Multi-level modelling : Multi-level modelling


Optimization model using LP : Optimization model using LP x1, x2,... vector of decision variables x xi = 0 : no, you do not do it xi = 1 : yes, you do it g1x1 + g2x2 + .. objective function gx --andgt; max a11x1 + a12x2 + .. andlt;b1 constraints Ax andlt; b a21x1 + a22x2 + .. andlt;b2


Non-linear programming : Non-linear programming non-linear constraints and/or non-linear objective optimality not guaranteed (lowest point potato field?) if optimality is guaranteed, then you can probably do it with LP (piece-wise linear)


Piece-wise linear yield function (convex) : Piece-wise linear yield function (convex)


Non-linear programming (ctd) : Non-linear programming (ctd) non-linear constraints and/or non-linear objective optimality not guaranteed (lowest point potato field?) if optimality is guaranteed, then you can probably do it with LP (piece-wise linear) if not guaranteed, then with integer programming you can construct non-linear functions using special sets


Use of special sets for constructing non-convex piece-wise linear functions : Use of special sets for constructing non-convex piece-wise linear functions


Approximation of quantity*quality : Approximation of quantity*quality (a+ x1)*(b+x2)  ab + ax2 + bx1


Influence matrix approach : Influence matrix approach


Building of simplified groundmodel : Building of simplified groundmodel Boundary condition of nature area in terms of Mean Spring Watertable MSW Mean Lowest Watertable MLW seepage that reaches the rootzone


Analytical solution for spatial interaction : Analytical solution for spatial interaction steady-state homogeneous geohydrology radial flow analytical solution (Groenendijk) Unit rise of head 0 Calculated effect j i 1


‘Walking’ measure : ‘Walking’ measure Influence matrix IM for spatial interaction through groundwater Bovenaanzicht Modelcel (i) j i j a(i)/p(j) a(i)/p(j) IM =


Combination with simulation model : Combination with simulation model k 1 k 2 2) grondwaterstand veranderingen 3) superpositie effecten op stijghoogten 5) grondwaterstand- veranderingen 4) stijghoogte- veranderingen 1) maatregelen landbouwgebied 6) effecten op natuurgebied Sensitivity analyses with SIMGRO (uniform measure) 2) MHW, MSW, MLW (phreatic level agricultural land) 4) MSWa en MLWa (aquifer under nature area)


Regression model MSWa (1) : Regression model MSWa (1) MSWa = fMSW · [IM] •  MSW


Regressiemodel MSWa (2) : Regressiemodel MSWa (2) MSWa = fMSW · [IM] • MSW  MSWa = fMSW · [IM] • MSW + fMHW · [IM] • MHW


Embedding approach using mixing cells : Embedding approach using mixing cells


Software : Software Xpress package of DASH interior point algorithm (not ‘Simplex') integer extensions (also binary variables) use of special sets for nonlinear functions implemented with integer variables


Pilot study/methodology : Pilot study/ methodology


What are we talking about ? : What are we talking about ? 1. Problem definition


Pilot area Beerze & Reusel : Pilot area Beerze andamp; Reusel


What are the stakeholder objectives ? : What are the stakeholder objectives ? 1. Problem definition 2. Objectives - stakeholders - authorities


Objectives : Objectives reduce flood risk / climate change reduce desiccation of nature areas reduce nitrogen and phosphorous loading on groundwater andamp; surface water minimize loss of income from agriculture


Where are we now ? : Where are we now ? 1. Problem definition 2. Objectives - authorities - stakeholders 3. Actual situation - now


Situation Now land use : Situation Now land use


AlterrAqua: GIS-shell for regional hydrology : AlterrAqua: GIS-shell for regional hydrology waterways subcatchments DTM sewerage systems culverts weirs Land use top10 vector


Metamodel for leaching of nutrients : Metamodel for leaching of nutrients Pload = f(Soiltype, Landuse, P-surplus, MHW)


Situation NowNitrate concentration(in the long-term,after endlessly repeating manuring) : Situation Now Nitrate concentration (in the long-term, after endlessly repeating manuring)


Catchment accumulation of NO3-N loadingon surface water : Catchment accumulation of NO3-N loading on surface water


Situation Now : N-loading on surface waternitrogen surplus : Situation Now : N-loading on surface water nitrogen surplus


Where are we heading ? : Where are we heading ? 1. Problem definition 2. Objectives - authorities - stakeholders 3. Actual situation - now - autonomous developments


Autonomous developments + climate scenario : Autonomous developments + climate scenario Discharge (m3/s) Situation Now Pwinter +17% Autonomous dev.


Autonomous developments: drainage & nature : Autonomous developments: drainage andamp; nature Current Situation Autonomous development


What should we focus on ? : What should we focus on ? 1. Problem definition 2. Objectives - authorities - stakeholders 3. Actual situation - now - autonomous developments compare 4. Focal points


What are the options ? : What are the options ? 1. Problem definition 2. Objectives - authorities - stakeholders 3. Actual situation - now - autonomous developments compare 4. Focal points 5. Measures (options)


Measures(options) : Measures (options) land use water management


What is the best strategy ? : What is the best strategy ? 1. Problem definition 2. Objectives - authorities - stakeholders 3. Actual situation - now - autonomous developments compare 4. Focal points 6. Strategies 5. Measures (options)


Planning with water, ‘inverse approach’ : Planning with water, ‘inverse approach’ optimization Stakeholders: targets on objectives options for measures measures communication


Integration with agricultural model DRAM : Integration with agricultural model DRAM


Contribution to peak flow, per subcatchment : Contribution to peak flow, per subcatchment


Contribution topeak flow in reference run : Contribution to peak flow in reference run


Optimisation-model (Beerze-Reusel) : Optimisation-model (Beerze-Reusel) 60 000 constraints 200 000 continuous decision variables 2 million non-zero coefficients in de matrix CPU-time ~0.5 hour on a P4-2.4


Strategy 1 : flood risk  : Strategy 1 : flood risk  Discharge (m3/s) Situation Now Pwinter +17% Autonomous dev. Strategy 1


Strategy 1 (ctd) : generated pattern of measures : Strategy 1 (ctd) : generated pattern of measures


Strategy 2a:- desiccation  - option for new natural grasslands discourageddY = - 1.5 M€ /ydN = 64 ha : Strategy 2a: - desiccation  - option for new natural grasslands discouraged dY = - 1.5 M€ /y dN = 64 ha


Strategy 2b:- desiccation  - option for new natural grasslands encourageddY = - 0.7 M€ /ydN = 250 ha : Strategy 2b: - desiccation  - option for new natural grasslands encouraged dY = - 0.7 M€ /y dN = 250 ha


Strategy 3: combined targets of 1&2 dY=-3.3M€/y : Strategy 3: combined targets of 1andamp;2 dY=-3.3M€/y 3. flood risk  desiccation  1. flood risk 


Strategy 3 (ctd.) : field drainage : Strategy 3 (ctd.) : field drainage Field drainage 1. flood risk  3. flood risk  desiccation 


Landuse in strategy 2b : Landuse in strategy 2b 3. flood risk  desiccation  2b. desiccation 


Implicit conflict: flood risk  <> desiccation  : Implicit conflict: flood risk  andlt;andgt; desiccation 


Strategy 4:- flood risk  - desiccation  - N-loading SW  : Strategy 4: - flood risk  - desiccation  - N-loading SW  dY = - 17.5 M€ /y


N-loading on surface water generic measures <> optimisation : N-loading on surface water generic measures andlt;andgt; optimisation


Slide60 : 4a N-loading SW  via generic measure dY = -25 M€/j dY = -17 M€/ j 4 N-loading OW  via optimisation


Tradeoff curve for SW-objective and income : Tradeoff curve for SW-objective and income With (rood) and without (blue) transport via deep groundwater


Trade-of curves desiccation nature : Trade-of curves desiccation nature naturezones 1-20


Trade-off curves nature areas : Trade-off curves nature areas


Concluding remarks (modelling) : Concluding remarks (modelling) integrated modelling of hydrology, ecology, and economy combined use of simulation andamp; optimization turns the regional system ‘inside-out’ ideas for solutions, gives insight


Concluding remarks (stakeholder participation) : Concluding remarks (stakeholder participation) stakeholders must get really creative about options (including multifunctional forms of landuse) protocol for interaction with stakeholders must be further developed (in the form of a game?)


Increased adaptive capacity through risk diversification : Increased adaptive capacity through risk diversification


Slide67 : GHG veldinventarisatie SIMGRO vóór calibratie na downscaling (25*25m) Afvoer (l/s)


Slide68 : GHG veldinventarisatie SIMGRO na calibratie en downscaling (25*25m) Afvoer (l/s)


Calibration factors for MSWa : Calibration factors for MSWa fGHG (-) fGVG (-)


Verification with SIMGRO (1) : Verification with SIMGRO (1)


Verification with SIMGRO (2) : Verification with SIMGRO (2)