Combining hydrologic modeling and economic factors to optimize water management for a vietnamese wetland system

41 2 DEVELOPMENT OF A WETLAND HYDROLOGIC MODEL FOR WATER MANAGEMENT CONSIDERING WATER ELEVATIONS TARGETS IN TRAM CHIM NATIONAL PARK, VIETNAM .... 68 3 MODELING DECISION-MAKING REGARDING

COMBINING HYDROLOGIC MODELING AND ECONOMIC FACTORS TO OPTIMIZE

WATER MANAGEMENT FOR A VIETNAMESE WETLAND SYSTEM

By

TANH T.N NGUYEN

A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE

UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT

OF THE REQUIREMENTS FOR THE DEGREE OF

DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA

2014

© 2014 Tanh T N Nguyen

To my wife Đoàn Thị Minh Nguyệt, son Nguyễn Đoàn Hữu Phúc, parents Trần Thị

Du-Nguyễn Ngọc Hân, and parents-in-law Đoàn Hữu Lực-Từ Thị Bạch Tuyết

ACKNOWLEDGEMENTS

I give special thanks to Dr Kati W Migliaccio for giving me the opportunity to study at the University of Florida Her coaching, advising, and support are lessons for

my career and my life I always remember the time I contacted her from Vietnam in

2010 Without her, I would not have my today

I would like to thank my PhD committee members for helping me sharpen my skills to complete this dissertation I really felt a little worried for selecting the hydro-economic coupling topic for my study Dr Christopher J Martinez helped me much in modeling working with data limitation The advice of Dr Mark W Clark in wetland

guided me to concentrate on wetland issues Support of Dr John J Sansalone led me

to the right ways to achieve my study objectives Dr Edward A Evans made my life easier as working on environmental economics and happier for his leading events in Homestead, particularly Friday night karaoke that helped me recovered and have more energy to work for my dissertation

I would like to thank Dr Dorota Z Haman She gave me the best chance for

studying in a warm environment in the ABE department I am proud to be part of the Department of Agricultural and Biological Engineering

Finally, I would like to give special thanks to organizations supporting me very much for my study: An Giang University (Vietnam), Department of Agricultural and

Biological Engineering and Tropical Research and Education Center (University of

Florida, US) for PhD research assistantship, McNair PhD Bostick Scholarship, Tram Chim National Park Authorities (Vietnam), Vietnam’s PhD Fellowship Program, and World Bank Robert S McNamara PhD Research Fellowship

TABLE OF CONTENTS

page

ACKNOWLEDGEMENTS 4

LIST OF TABLES 7

LIST OF FIGURES 9

LIST OF ABBREVIATIONS 10

ABSTRACT 12

CHAPTER 1 INTRODUCTION 14

Motivation 14

Objectives 17

Research Contribution 17

Reviews of Hydro-economic Modeling 18

Hydrologic Models 18

Rainfall 19

Evapotranspiration 20

Infiltration 24

Water routing simulation 25

Wetland Hydrologic Models 27

Wetland Hydro-Economic Modeling 31

Economic valuation 31

Economic analysis 34

Hydro-economic models 37

Model Sensitivity Analysis, Calibration and Validation 41

2 DEVELOPMENT OF A WETLAND HYDROLOGIC MODEL FOR WATER MANAGEMENT CONSIDERING WATER ELEVATIONS TARGETS IN TRAM CHIM NATIONAL PARK, VIETNAM 43

Background 43

Methods 47

Study Site 47

Measured Data Evaluation 50

Model Development 51

Model design 51

Model evaluation 55

Scenario Optimization 57

Results and Discussions 57

Measured Data Assessment 57

Wetland Hydrologic Model Assessment 61

Optimization of Water Scenarios 63

Assessment of the Targets and Water Control Protocols 67

Summaries 68

3 MODELING DECISION-MAKING REGARDING WETLAND SERVICES FOR WETLAND MANAGEMENT IN TRAM CHIM NATIONAL PARK, VIETNAM 70

Problems 70

Methodology 76

Study Site 76

Conceptual Study Approach 78

Selection of stakeholders (step 1a) 78

Analysis of wetland management priorities (step 2) 79

Selection of key wetland management priorities (step 3) 79

Valuation of the key priorities (step 4) 80

Analysis of scenarios (step 5) 84

Initiation of decision conditions (step 1b) 84

Programming Languages 84

Results of Application of the Study Framework for Tram Chim National Park 84

Decision Groups and Wetland Priorities 84

Analysis of Wetland Management Priorities 87

Fish exploitation 87

Tourism 90

Management costs 91

Scenario Analysis 92

Summaries 94

4 COUPLING HYDROLOGIC AND ECONOMIC MODELING FOR WETLAND MANAGEMENT IN TRAM CHIM NATIONAL PARK, VIETNAM 96

Rationale 96

Methodologies 99

Study Site 99

Conceptual Study Approach 100

Model Design 101

Bridging HMTC and EMTC 105

Coupling Criteria and Scenario Analysis 105

Results 106

Bridging between HMTC and EMTC 106

HEMTC Limits and Boundaries 106

Optimal Scenarios for Four Wetland Zones 108

Summaries 112

5 CONCLUSIONS 113

LIST OF REFERENCES 115

BIOGRAPHICAL SKETCH 139

LIST OF TABLES

1-1 Rainfall estimation methods 19

1-2 Evapotranspiration estimation methods with necessary parameters 22

1-3 Wetland surface/groundwater interaction models 28

1-4 Wetlands models with no groundwater component 29

1-5 An assessment of wetland hydrologic models 30

1-6 Use of economic analysis in assessing wetland goods and services 34

1-7 Descriptions of some hydro-economic models 40

2-1 Suggested monthly water elevations (m, above mean sea level) adapted from Van Ni et al (2006) 45

2-2 An assessment of wetland hydrologic models from literature 46

2-3 Properties of wetland hydrologic model components 52

2-4 Classification of water elevations used for modeling 60

2-5 Goodness-of-fit values for simulated water elevation for zones A1, A2, A4, and A5 61

2-6 Optimal wetland water elevations for zone A1 and MWBP targets for 2007-2011 (m) 64

2-7 Optimized wetland water elevations for zone A2 and MWBP targets for 2007-2011 (m) 64

2-8 Optimized wetland water elevations for zone A4 and MWBP targets for 2007-2011 (m) 65

2-9 Optimized wetland water elevations for zone A5 and MWBP targets for 2007-2011 (m) 66

3-1 Net income and tickets of fish exploitation per season (September-December) in 2011 ($ unit: USD, N= 46) 88

3-2 Parameter estimates for zone A1 (n=194, AIC= 421.2) 90

3-3 Willingness to pay in addition to entrance fees (WTPa) by wetland zones in

2013 ($ unit: USD) 91

3-4 Annual management cost distribution for four wetland zones ($ unit: 1000 USD) 91

3-5 Comparisons of annual benefits by zones ($ unit: USD) 93

3-6 Relationship between WTP of fishing types and desired plant communities ($ unit: USD) 93

4-1 Properties of hydrologic model components 103

4-2 Design properties for economic model 104

4-3 Calendar for services and water control (X: active time) 107

4-4 Annual fish exploitation limits in wetland zones 108

4-5 Annual management costs, total net benefits, and Park’s net benefits regarding double optimization ($ unit: 1000 USD) 109

4-6 Redistribution of fishermen for double optimization at zone and Park levels ($ unit: 1000 USD) 110

LIST OF FIGURES

1-1 Map of Tram Chim National Park with canals, wetland zones (A1, A2, A4,

and A5), water gates (C1, C2, C3, C4, C5, C6, C7, and C8), and berms 162-1 Map of Tram Chim National Park with canals, wetland zones (A1, A2, A4,

and A5), water gates (C1, C2, C3, C4, C5, C6, C7, and C8), and berms 452-2 Measured water elevation medians with standard deviation and MWBP

targets for A1 and A2 for 2007-2011 492-3 Measured water elevation medians with standard deviation and MWBP

targets for A4 and A5 for 2007-2011 492-4 Water balance diagram with water elevation (the model output) and

inflows/outflows used in model development 522-5 Water elevation spectra in canal at C1 and C7 sites for 2007-2011 582-6 Water elevation spectra at HMC site for 2007-2011 (maximum, minimum,

and average are daily water elevation maximum, minimum, and average of

minimum and maximum, respectively) 592-7 Model of water lost and filled in canals outside the Park 612-8 Simulated water elevation medians with standard deviation and MWBP

targets for A1 and A2 for 2007-2011 622-9 Simulated water elevation medians with standard deviation and MWBP

targets for A4 and A5 for 2007-2011 623-1 Map of Tram Chim National Park, Vietnam, with wetland zones 773-2 Decision-making framework using MCDA based on net benefit 783-3 Scores for priorities of the decision-makers at Tram Chim National Park,

using analytic hierarchy process and pairwise comparison (38 pairs from 7

priorities for each person) 853-4 WTP of fishermen for preferred plant communities in 2011 (n=46) 894-1 Tram Chim National Park map with canals, water gates (C1, C2, C3, C4, C5,

C6, C7, and C8), berms, and wetland zones (A1, A2, A4, and A5), which was adapted from Nguyen & Migliaccio (2014) 994-2 Framework for coupling simulation of hydrologic and economic models to

search for optimal scenarios using anneal scheduling 102

LIST OF ABBREVIATIONS

Analytic hierarchy process Chief of Office for Science of Tram Chim National Park Consumer surplus/access fees

Contingent valuation method Absolute deviation between observed and simulated water elevations

Economic model for Tram Chim National Park Potential evapotranspiration

Full water control system using water gates Hydro-economic model

Dong Thap Province Hydro-Metereological Center at Tram Chim station

Hydrologic model for Tram Chim National Park Multi-criteria decision analysis

More natural water control system using berms Nash-Sutcliffe efficiency

Tram Chim National Park Percent bias

Director of Tram Chim National Park Net benefit for the Park only

Root mean square error

Revealed preference method RMSE-observations standard deviation ratio Stated preference method

Travel cost method Techinical Group of Tram Chim National Park Total economic value

Total net benefit including revenue for the Park and fishermen's income

Willingness to pay Additional amount tourists were willing to pay

Abstract of Dissertation Presented to the Graduate School

of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy COMBINING HYDROLOGIC MODELING AND ECONOMIC FACTORS TO OPTIMIZE

WATER MANAGEMENT FOR A VIETNAMESE WETLAND SYSTEM

By Tanh T N Nguyen December 2014

Chair: Kati W Migliaccio

Major: Agricultural and Biological Engineering

Responding to the demand of a comprehensive coupling framework for scenario analysis regarding multiple objectives for wetland, we developed and tested our

coupling framework for applicability using a case study in the Tram Chim National Park applying full water control (FC) using water gates (zones A1 and A2) and more-natural water control (NC) using berms (zones A4 and A5) The study objective is identifying scenarios meeting double optimization criteria including hydrologic optimization

(monthly water elevation as targets suggested by the Mekong Wetland Biodiversity Programme) and economic optimization (net benefits of wetland management is equal

or greater than 0) Methods consisted of (1) development of a hydro-economic model (HEMTC) using R and Java languages, which coupled water balance model (HMTC) (applying data screening with spectral analysis and filling-in approach and using

goodness-of-fit indicators and a categorical comparison for model evaluation) and an economic model EMTC (including decision [analytic hierarchy process], tourism [travel costs and contingent valuation with Poisson regression], and fish exploitation

[contingent valuation and consumption approach]); (2) multi-criteria decision analysis (MCDA) for scenario analysis; and (3) anneal scheduling for scenario optimization HEMTC simulation indicated that water elevation in zone A1 met the targets but that other zones did not Economic benefits of wetland services for all zones were

US$159/year/person for fish exploitation and US$11/person of consumer surplus for tourism Net benefit for zone A1 was the highest among wetland zones Scenario

analysis using hydro-economic coupling simulation showed that water management meeting double optimization criteria required: no change of water control system and a fishing permission with 64 fishermen for zone A1; additionally opening water gate C5 in February and May and allowing 171 fishermen for fish exploitation for zone A2;

replacement of levees and 142 fishermen permitted for fishing for zone A4; and

replacement of levees combining with 109 fishermen allowed for fish exploitation for zone A5 This study suggested a consideration at zone level for hydrologic optimization while a Park level was selected for economic optimization This study showed the

applicability of our coupling framework for multiple objective studies, where four wetland zones were optimized in regards to hydrology and economics

CHAPTER 1 INTRODUCTION

Motivation

While hydrologic changes may impact goods and services of a wetland and thus

have economic consequences, most hydrologic modeling efforts on wetlands do not

include values of ecosystem goods and services in their evaluation of different

scenarios Determination of a truly optimized management scenario using wetland

models requires the integration of economic value assessment with wetland hydrologic

modeling particularly when multiple wetland uses are being considered Current wetland

models are complex and may have been developed for a specific application; their use

may be constrained by code/software availability, model flexibility, applicability with

limited data, and capacity of performing economic aspects For wetlands in developing

countries such as Tram Chim National Park in Vietnam with complex water

management objectives and limited data, available models are not viable and therefore

a new integrated wetland model that includes optimization of water management

objectives with limited data and economic assessment is needed

The Tram Chim National Park (Park), 7,500 ha, is part of the Mekong River Delta

of Vietnam (Figure 1-1) History of wetland conservation in the Park is linked with

hydrologic changes (van der Schans, 2006) Early settlement in the location was

initiated between 1800-1950 along with flora and fauna exploitation (i.e fish, bird, and

edible plants) The wetland was drained in 1980s due to economic renovation promoting

rice production Low levees were later constructed to hold water in dry season to

prevent fires in the Melaleuca trees and restore native flora Water gates were

developed in 1990’s to control water for Sarus Crane protection

The Park was certified as “Wetlands of International Importance” by the Ramsar Convention Wetlands (Ramsar) in 2012 and characterized by seasonal appearance of

the Sarus Crane (Grus antigone) which is listed in the International Union for

Conservation of Nature (IUCN) Red List of Threaten Species (Do and Bennett, 2009; Pilgrim and Tu, 2007) This species visits the Park for a short time during dry season Thus, water management has primarily aimed at protecting food areas for the Crane Over time, other water purposes were added to the Park’s strategic management As a result, four major water management objectives exist First, efforts for the Crane

protection resulted in construction of a system of high soil-levee around the Park to control water inflow and outflow by water gates (Smardon, 2009; Thanh, 2003) Water inflows are from rainfall and flood during an annual flood season (around 3 months) After floodwater reaches its peak and recedes, water gates are closed to hold water in the wetland The major purpose of inflow and outflow control is to store water in the Park at appropriate levels for maintaining food areas for the Crane Second, the Park provides habitat for aqua-resources, e.g., fish and edible plants, that are used by poor local residents for their daily food and income (Maltby et al., 2006; Shepherd, 2008; Thanh, 2003) Alteration of water operation may cause variation in the availability of fish and edible plants which impacts their daily sustainability Third, the Park is home to the

Melaleuca trees (Melaleuca cajuputi), a native species, classified as forest trees Under

the Vietnam forest protection law, the Park is under strict fire control, thus park

managers maintain these areas with extra water to prevent fire (Van Ni et al., 2006) Fourth, water control is needed to protect biodiversity (English and LaRoche, 1997; van der Schans and Thien, 2008) The water level in the Park is a direct factor in meeting

each of these objectives; however, the optimum water level for each objective often differs creating a conflict, which suggests an adaptive and multi-objective water

management strategy is needed

Figure 1-1 Map of Tram Chim National Park with canals, wetland zones (A1, A2, A4,

and A5), water gates (C1, C2, C3, C4, C5, C6, C7, and C8), and berms The Park Authority has experimented with a full control system with water gates that control water into and out of the Park (FC) and a more-natural system where water flows over berms into and out of the Park (NC) (Figure 1-1) FC includes zones A1 (10o42’49” N – 105o30’12” E; 4,942.8 ha) and A2 (10o42’06” N – 105o35’10” E; 1,122.7 ha) and NC consists of zones A4 (10o33’48” N – 105o35’44” E; 731.9 ha) and A5 (A5:

10o45’44” N – 105o29’54” E; 440.5 ha) However, neither strategy has been able to meet their water management objectives Another aspect of Park water management that has not been considered is the economics By adding economics, a more

realistically achievable water management scenario may be obtained

Objectives

The following objectives were identified for developing a conceptual model

coupling framework that applies theoretical hydro-economic modeling for use in wetland systems with complex water use issues:

1) Develop a hydrologic model HMTC (water balance) to predict water level variation inside the Park at 4 locations with satisfactory goodness-of-fit and categorical comparison

2) Design an economic model EMTC (water elevation decision, edible plants and fish exploitation, tourism, and net benefit) to assess decisions

regarding wetland services

3) Couple HMTC and EMTC to create HEMTC (hydro-economic model for Tram Chim National Park) to simulate optimal decisions for wetland

management considering hydrology and economics

Research Contribution

This study shows the adaptation of applying hydrological and economic modeling

to solve real-world and multi-faceted wetland water conflicts Modeling protocols from linking economics to hydrology for both model design and simulation were derived The HEMTC in this research not only provides a scientific base of hydro-economic modeling for wetlands but also provides a real-world test to identify an optimized solution for surface water management in the Tram Chim National Park, an international Ramsar wetland protecting biodiversity of the world The integrated hydro-economic modeling framework derived from this study may be applied to wetland water management

systems worldwide

Reviews of Hydro-economic Modeling

Hydrologic models and economic models have a particular structure, function, operation, and application Their differences complicate their combined application Thus, review of their properties and applicability is useful for identifying possible gaps and commonalities for integration

Hydrologic Models

Hydrologic models simulate a defined hydrologic system in a dynamic way so that different alternatives may be investigated Hydrologic models typically include various water storage components (e.g., surface water, groundwater, and soil water) and

hydrologic processes (e.g., infiltration, stream flow, runoff, and evapotranspiration) Wetland plant communities rely upon water elevation and their existence is a function of hydrologic processes

For wetland hydrologic studies, water balance calculations are fundamental and commonly represented as

considering water targets

Researchers have applied the water balance approach successfully to evaluate wetland systems Hammer and Kadlec (1986) estimated inflows (from precipitation, stream flows, and groundwater discharge) and outflows (from evapotranspiration, groundwater recharge, runoff, and pumped conditions); Croke et al (2000) simulated a water budget using the rainfall-runoff model MESSARA; and Kunkel and Wendland (2002) used the GROWA98 model for a water balance analysis at a basin scale

Rainfall

Rainfall (precipitation) is one of the major water inputs of a wetland and often a source of uncertainty in hydrologic modeling (McMillan et al., 2011; Renard et al., 2010; Salamon and Feyen, 2009) Rainfall uncertainty results from spatial variation of rainfall Rainfall point data should be measured using a tipping bucket or other rainfall

measurement device; these point measurements may be used to estimate rainfall at other locations using rainfall estimation methods (Table 1-1)

Table 1-1 Rainfall estimation methods

and Singh, 2005) Radar Uses radar to estimate rainfall but results depend upon evaporation

(Dingman, 2008) and radar-resolution (Maidment, 1993) Thiessen

a R R

A



 

(1-2)

where R is the mean of rainfall, a i is the area of ith gauge sub-region, R i is the rainfall at

ith gauge, and is the total area of the sub-regions While the Thiessen polygon is based on weighting, the isohyetal method uses isohyetal maps for rainfall estimation (Fiedler, 2003) Isohyetal maps have been developed for the entire US using the

Precipitation-Elevation Regressions on Independent Slopes Model which was based on

a digital elevation model (Daly et al., 1994) The isohyetal method provides greater accuracy in rainfall estimation as compared to the Thiessen polygon (Fiedler, 2003; Warwick and Haness, 1994)

Evapotranspiration

Evapotranspiration (ET) partitions to evaporation (of liquid water from the surface

or soil) and transpiration (release of water vapor from plants) Evaporation may be measured or estimated separately from transpiration, particularly for water bodies The pan evaporation method for a water body is used by the National Climate Data Center

of the U.S to calculate evaporation This pan method uses a standardized Weather Bureau class A pan which is an iron tank with 1.21 m of diameter and 0.255 m of depth that is mounted 0.3048 m above the ground (Bedient et al., 2008) The tank is filled with 0.2032 m water deep Evaporation is computed daily based on the difference of

observed water levels in the pan

The Dalton-type method may also be used for evaporation estimation which is based on vapor pressure (Brutsaert, 1982; Dingman, 2008):

(es a)( )

where e s is the saturation vapor pressure, e a the vapor pressure at fixed level above

water surface, v is wind speed, and m and n are empirical constants The Dalton-type, however, requires a selection of a proper wind speed height to measure wind speed v

(Bedient et al., 2008) Also, e a is difficult to be obtained due to limited temperature data above water surfaces (Dingman, 2008) The Dalton-type equation may be reduced to

simplify the evaporation estimation, e.g., m is set to zero and the pressure and wind

speed are measured at 2 m above water (Harbeck and Meyers, 1970)

Bedient et al (2008) suggested that the energy budget method was more accurate than the Dalton-type method for lake evaporation calculation:

where Ra N is the net radiation absorbed by water body, Ra h is the sensible heat

transfer, Ra e is the energy used for evaporation, Ra ɵ is the increase in energy stored in

the water body, and Ra v is the advected energy of inflow and outflow Compared to the pan method and the Dalton-type method, the energy budget method is more

complicated due to its dependence on radiation data

Evaporation for a water body may be also estimated by the Penman method that includes mass transfer and the energy budget The Penman does not require data of soil and water surface temperatures (Bedient et al., 2008; Dingman, 2008) The

Penman equation for open water is (Penman, 1948):

the air, E a is the air “drying powder”, a+bu is the empirical constants per unit of

pressure, e sa is the saturation vapor pressure at temperature of the air, and e a is the actual vapor pressure in air relative humidity

Table 1-2 Evapotranspiration estimation methods with necessary parameters

Methods/

Sources

Parameters

Application Crop Hu NR SHF SR T0 WS Others

Carlson et al

(1995)

difference vegetation index

Short and tall crop

SR is solar radiation, T0 is temperature, WS is wind speed, Others is other parameters, Application is applicability of the models for particular conditions, and General is the application in general

Evaporation may be combined with a transpiration rate to compute ET ET may be alternatively estimated using weather-based equations Various equations are available

to estimate reference ET (ETo) or potential ET (ETp) (Table 1-2) The Holdridge (1959)

is the simplest method as it is based solely on temperature Both the Hargreaves and Samani (1985) and Turc (1961) methods require less parameters to estimate ET than most other methods Although the Hargreaves’s method may adapt well in different climate conditions (Weiland, 2011), it is best suited in dry climates The Turc has been proven as applicable for areas with humidity levels of 50% or greater and predictable of

ETp values precisely similar to Penman-Monteith methods (Jacobs et al., 2001; Jensen

et al., 1990; Nandagiri and Kovoor, 2006; Trajković and Stojnić, 2007) The Turc

ETp data:

1/2 2

1

a

p

R ET

R ET

The combination of Turc method and Pike method (Turc-Pike methods) has been

applied for ET a estimation for tropical regions such as South America (Yates, 1997),

Malawi (Pike, 1964), and Honduras (Alvarez et al., 2004) The Turc equation has been

used successfully in water balance analysis to determine parameters controlling

impacts of climate change on catchments (Arnell, 1992); assessing influence of

seasonal precipitation anomaly on annual runoff (Milly, 2002); and studying impacts of climate change on water balance (Yates, 1997) The advantage of this approach is that

it does not require a crop coefficient Determination of crop coefficients may be very difficult especially for mixed plant and land cover systems

Infiltration

Infiltration, the movement of water into the soil profile from the ground surface, may be measured in the field using a ring infiltrometer (Maidment, 1993) Different methods for infiltration estimation have also been documented One method is the Horton (1940):

 0 

kt

where i is the infiltration capacity, i c the final capacity, i 0 is the initial infiltration capacity,

and k is the empirical constant The infiltration capacity, however, decreases with time

regardless of the amount of water for infiltration and rainfall intensity excess by the

Horton equation (Bedient et al., 2008) In addition, obtaining proper values of k for

calculation is difficult

The Green-Ampt method for infiltration estimation was found to be more accuracy than the Horton method in wetland research applications (Fiedler and Ramirez, 2000; Koob et al., 1999) Green-Ampt is based on Darcy’s Law and mass conservation

(Bedient et al., 2008) The time step for infiltration simulation in the Green-Ampt method

is hourly The Green-Ampt equation is:

2K sΨt

L

η

where in is the infiltration rate, K s is the saturated hydraulic conductivity, η is the

porosity, L is the volume of infiltration down to the depth, ψ is the wetting front suction head, and is the time Estimating the Green-Ampt parameters (Equations 1-11 and 1-12) are difficult to quantify and thus the parameters from Rawls et al (1983) based on soil type data are often used

Water routing simulation

Water routing refers to the simulation of surface water movement The purpose of routing is to simulate storage (S), inflow (I), and outflow (O) as needed The general form of the I/O equation is (Maidment, 1993):

where S is storage and t is time Hydrologic routing methods applicable for wetland

systems include the modified Puls, level-pool method (storage as a nonlinear function of

O), the Muskingum method (storage as a linear function of I and O), reservoir models (storage as a linear function of O), and kinematic channel method

The modified Puls routing outflow over a weir or horizontal weir flow is calculated using the following equation

3 2

The level-pool method, levee overtopping, computes the outflow using the

following equations (Maidment, 1993):

McCarthy and US Army Corps of Engineers (1939) developed the Muskingum:

where a, b, m, and n are constants Assuming flow/ storage are related to depth (m/n =

1 and b/a = K [travel time]), the linear Muskingum equation is (Bedient et al., 2008):

1 

where x is weighting factor (0 to 0.5) The Muskingum could be applied for flood routing

for urban storm water treatment by a catchment wetland (Wong et al., 2002) For

reservoirs having S depending only on outflow and x = 0, reservoirs models may be

derived from the Muskingum (Equation 1-18) to become:

The reservoir models are applicable for water storage evaluations (Arora and Boer, 1999); thus they are applicable for wetlands that function as water storage systems

Water transport in wetlands may be controlled by channels and routed using kinematic channel method Parameters required for the application of kinematic channel method include cross-sectional dimensions, slope, length, shape, and a Manning value The kinematic channel equation is (Bedient et al., 2008):

c

m

c c

where Q is inflow, O is outflow, α c and m c are the kinematic wave parameters (Manning

value) for a specific channel, and A c is the cross-sectional area of the channel αc and

m c may be obtained from US Hydrologic Engineering Center (Hydrologic Engineering Center, 1993)

Wetland Hydrologic Models

Wetland hydrologic models are developed to investigate wetland issues, including surface and groundwater interactions (Table 1-3) and water balance simulations (Table 1-4) In order to support particular study purposes, wetland hydrologic models are often designed with specific applications for desired study sites The selection of patterns for design for wetland hydrology depends upon modeling expertise of researchers and thus they are very diverse For example, researchers used various software to develop the models, e.g., Zhang and Mitsch (2005) and Twilley and Chen (1998) applied STELLA software and Vining (2007) used the Microsoft Excel software Developers often

designed their models for specific research purposes and have not maintained or

updated their code for additional applications such as the models of Hammer and

Kadlec (1986), Mansell et al (2000), and Kazezy et al (2007) Other models, however, are supported for continued use, e.g., the HEC-HMS (Ta and Brignal, 1998),

DRAINMOD (Workman, 1994), and MIKE-SHE (Singh and Yadava, 2003)

Table 1-3 Wetland surface/groundwater interaction models

Models Distributed

model

DRAINMOD MIKE SHE STELLA-

based

WETLANDS WETSAND CASC2D

Description Hammer and

Kadlec (1986)

Dept of Biological

& Agricultural Engineering, North Carolina State

University

DHI group ISSEE

company

Mansell et al.(2000)

North Carolina Agricultural Experiment and Skaggs (1980)

Water Environment Health (DHI)

Robert Webb of Australia

causing high

sensitivity

N/A Soil hydraulic

(Workman, 1994)

discharge, peak flow Publication Hammer and

Kadlec (1986)

McCarthy et al

(1992), Skaggs (2008)

Hughes and Liu (2008) Singh and Yadava (2003)

Zhang and Mitsch (2005), Twilley and Chen (1998)

Software

availability

Propriety Public Propriety Propriety Propriety Propriety Propriety

Types of

wetlands/

location

Peat land, Michigan

Wetland, North Carolina

Diverse Floodplain of

the Olentangy River, Ohio and Rookery Bay, Florida

Cypress pond, Florida

The Duke University restored wetland, North Carolina

Diverse

Application Investigate

hydrologic processes

Estimate drainage Simulate water

flow

Investigate hydrologic processes

Investigate hydrologic processes

Investigate surface flow and solute transport

Predict runoff and stream- flow

Note: N/A is not available

Table 1-4 Wetlands models with no groundwater component.

Models Simple

Refuge Stage Model

WASA Numerical

EXCEL model

HEC-HMS WAWAHAMO

(Germany)

ALSM-DEM GIS-based GTOPO30-

MOD12 Q1

Kerski (2012) Developer Meselhe et

N/A N/A N/A Water

storage LAI

Publication Meselhe et

al (2010)

Güntner et al.(2004)

Vining (2007) Maidment

(1993), Ta and Brignal (1998)

Zierl (2001) Kirk et al

(2004)

Brown et al (2010)

Shimizu (2010)

Research

purposes

Assess scenarios

Simulate runoff availability

Assess responses

of wetland

to hydrologic extremes

Simulate rainfall- runoff

Simulate water balance in

an entire forested area

Evaluate and compare several restoration alternatives

Analyze Water balance

Analyze water budget

Software

availability

Propriety Propriety Propriety Pubic Propriety Public Propriety Public

Wetland

types/

Location

Everglades marsh

Semi-arid Northeast

of Brazil

Fort Berthold Reservation

of North Dakota

Basins, wetlands, reservoirs

Swiss forest areas

Levy Prairie, Alachua County, Florida

Clay setting areas, Florida

Mekong River Basin forested areas

*Notes: WASA is model of water availability in semi-arid environment, ALSM is an airborne laser swath mapping, DEM is digital elevation model,

US Corps is US Army Corps of Engineers, and N/A is not available.

Wetland hydrologic models (Table 1-3 and Table 1-4) have not been used to investigate wetland ecosystem goods and services while these values may affect the decision-making process The identified models were developed primarily for research

in the US and Europe For these reasons, application of these wetland models for wetland research in tropical and less developed countries (where available data may be limited) may not be viable Thus, wetland models need to be evaluated to determine their appropriate application in a system such as Tram Chim National Park considering the following criteria: availability of software or source code (AVA); flexibility in

processes simulated (FLEX); verified and in peer reviewed literature (VER); applicable with limited data available for the system (APL); and able to evaluate hydrology and economic aspects of a wetland (EVA) Reviews of recent wetland models show no model met all criteria and the criterion EVA; and, only four out of 15 models satisfy the criterion APL (Table 1-5)

Table 1-5 An assessment of wetland hydrologic models

Wetland Hydro-Economic Modeling

Hydrological models provide information on how different scenarios are predicted

to alter the hydrology but they provide no link between the hydrological outcome and the economic impact As many decisions need to consider both aspects to identify a viable solution, there is a need to integrate these two factors To include economics, the value

of different wetland ecosystem goods and services must be quantified Values are determined for these goods and services by conducting an economic valuation These values may then be considered more holistically using analytical methods

Economic valuation

Economic valuation, fundamentally, is an estimation of the net value of a good to the public after subtracting market cost for protection of that good (Thurston et al., 2009) The economic valuation provides a quantitative method for comparing existing environmental services with hypothetical environmental services to balance decision-making (de Groot et al., 2006; Freeman, 1993) Wetlands have environmental services, e.g., tourism, biodiversity protection, and local people’s income that should be

economically valued to support decision-making

The valuation of ecosystem goods and services often uses surveys (Mäler and Vincent, 2005) Surveys may be used to assess stakeholders’ willingness to pay (WTP) for a particular ecosystem good or service (Carlsson et al., 2003); which is used to measure demand and value (Booker et al., 2012) WTP is applicable for estimating both use and nonuse economic values (Thurston et al., 2009) Use economic values in

wetlands may be the values from selling fish while the nonuse economic values are the values from ecosystem preservation For example, WTP helped revealed the perception

of local people placing economic values on protection of wildfowl habitat (Hammack and

Brown, 1974), removal of two dams on the Elwha River on the Olympic Peninsula in Washington State (Loomis, 1996), and protection of wetland species in the Everglades (Milon and Scrogin, 2006)

The valuation includes direct and indirect market valuation The direct valuation relates to market price, factor income, and public investments while the indirect

valuation considers avoided cost, replacement cost, mitigation or restoration cost as well as travel cost (de Groot et al., 2006) The indirect market valuation assesses

indirect values that people gain from goods or services For wetlands, market-based values often include fish values that may be obtained from the fish market, and indirect values may be determined by WTP investigation of wetland protection (Brouwer et al., 2009) The indirect and direct valuations based on monetary assessment are useful for comparing scenarios as well as values of wetland services, which include revealed preference method (RP) and stated preference method (SP) (Booker et al., 2012)

RP includes travels cost method, hedonic property value model, market method, and production approach The travel cost method, namely the nonmarket valuation technique (Bhat et al., 1998), may show how much people spend for travel to a

recreation area (Phaneuf and Smith, 2005) For example, the travel cost method may show costs for motor boating and waterskiing, fishing, and sightseeing in water-based eco-regions (Bhat et al., 1998) The hedonic property value model decomposes an item into its characteristics and measures values of each characteristic, e.g., values of ecosystem goods and services are estimated using a statistical relationship between characteristics of surface freshwater and prices of a good related to the water (Wilson and Carpenter, 1999) An example for using the hedonic models is an estimation of

amenity value of wetlands to nearby residential properties (Bin and Polasky, 2005; Mahan et al., 2000) The market method is the method that determines WTP for a service or good use using referred market prices, e.g., sale of wood in market The production approach investigating impacts of a service or good on economic outputs (Liu et al., 2010)

SP consists of direct value investigation based on proposed changes Some methods may be applicable for wetland research such as the conjoint analysis and the contingent valuation method (CVM) that values compensation for ecologic service change (Liu et al., 2010) The conjoint requires interviewees to rank their alternatives of preferences (Freeman, 1993) The conjoint may provide the information of an order of ranking of ecological conditions or values meaningful to specific groups of people For example, ranking of the WTP for stream clean-up using the conjoint analysis showed the group of people living in 50 miles had impacts on improvement of a stream (Farber and Griner, 2000) While the conjoint refers to ranking, the CVM calculates direct and indirect values of economic, and asks people to assign a value on a change (Van Den Bergh et al., 2004) Uncertainties of the CVM is due to the type of interview, provided scenarios, and how interviewers conduct interviews (Van Den Bergh et al., 2004) The CVM was applied to determine use and non-use values of wetland with WTP for

recreation, plants, flood protection, water supply, and wetland types protection (Stevens

et al., 1995) For example, the CVM was implemented to assess the WTP of residents

in Catawba River basin for a designated plan to protect the current water quality level (Kramer et al., 2009) that provided different considerations of policy makers, regulators and the residents considering costs and benefits associated with the protection of

ecosystems in the Catawba region In another case, Loomis (1996) applied CVM to estimate the WTP when considering removal of two dams on the Elwha River in

Washington State and restoring the ecosystem

Economic analysis

Economic valuation may provide information of wetlands’ values; however,

wetland decision making may not solely be based on the values from the valuation but may also need to consider factors related to a wetland on a contextual scale Cost-effectiveness (CEA) and Cost-benefit analysis (CBA) are economic analysis tools that value a choice or management decision from an economic perspective based on the valuation of goods and services While CEA refers to cost issues, CBA relates to both costs and benefits Applications of the valuation methods and economic analyses have been described in several studies (Table 1-6)

Table 1-6 Use of economic analysis in assessing wetland goods and services

Location

Hadejia-Nguru, Nigeria

Prairie potholes

Mangrove in Bintuni Bay

Louisiana coastal wetlands

Norfolk Broads/ East Anglia Method Partial

valuation to figure costs

to loss

Valuation of waterfowl and hunting

by CVM

Valuation of benefits of erosion control and biodiversity

CVM and travel cost method

CBA using CVM

Outcome Direct use

values through resources exploitation

Optimize numbers of pothole sites

Comparing scenarios of forest management

Total economic values for recreation

Ratio of benefit to cost for wetlands protection Problems Less water to

wetlands for dams/

irrigation

Impacts conversion wetlands to agriculture

Exploitation of charcoal, wood, fish ponds

Identification

of direct/

indirect uses

Impacts of flood protection

to wetlands

(1993)

Hammack and Brown (1974)

CEA is the least-cost approach that a project is considered as the most effective when it has the lowest investment cost (Thurston et al., 2009) The cost-effectiveness analysis equation is (Jacobsen, 2007):

AC CER

project CEA requires projects must have similar goals to be compared effectiveness (Levin and McEwan, 2000) that causes its disadvantages

Several recent worldwide studies applied CEA to simulate restoration scenarios and compared them through estimating costs The application of CEA was often for assessment of effectiveness of an investment or a change regarding costs For

example, CEA was used to assess effectiveness of changes for impacts on the river flow or on the resource (Berbel et al., 2011) In Florida, construction costs, outcomes, and conservation goals of different scenarios associated with water budget analysis were estimated to select an optimal long-term solution for wetland restoration using CEA (Kirk et al., 2004) CEA was also applied to evaluate service costs for wastewater treatment of semi-natural wetlands compared to traditional wastewater treatment plants

in Venice Lagoon watershed (Mannino et al., 2008) or effectiveness of wastewater filtration for phosphorus of wetlands compared to wastewater treatment plants in

Germany (Trepel, 2010)

CBA is widely used for water service valuation (Harou et al., 2009) and often used for decision-making analysis (Thurston et al., 2009) CBA refers to an assessment of alternatives with their monetarily valued costs and benefits (Levin and McEwan, 2000) Values of cost and benefit, especially marginal net benefit values, are useful for wetland management decisions (Barbier et al., 1997; Turner et al., 2000) The CBA equation is (Munasinghe, (1993):

1 t

B C NPV

compared to artificial fertilizers The benefits of future use values are more complex than the actual benefits due to limitation of knowledge of wetland valuation (Van Den Bergh et al., 2004)

A comprehensive CBA should include analyses of impacts on goods and services

of wetlands (Turner et al., 2000) Results from CBA and NPV provide values associated with different societal groups and economic values for water management decision making (Ward and Pulido-Velázquez, 2008) For example, CBA was used by George et

al (2011) to provide NPVs for changes in the Musi sub-basin in India to support wetland management decisions Several studies applied Bayesian networks for CBA for various purposes, e.g., pros and cons for nutrient abatement in integration water management

in the Morsa catchment (Barton et al., 2008) and uncertainty of water quality in

Tasmania, Australia (Kragt et al., 2011)

Hydro-economic models

Hydro-economic models link hydrologic and economic factors together using interrelationships between them The concept of combining economic and hydrologic models was reported by the Harvard Water Program (Maass, 1962) Costs and benefits may be integrated to hydrologic model to develop hydro-economic model (HEM) (Ward and Pulido-Velázquez, 2008) Common boundaries for the hydro-economic models are inflows and outflows (Letcher et al., 2007) HEM includes basic parameters such as hydrologic flows, water management infrastructure, economic water demands,

operating costs, and operating rules (Lund et al., 1999) Johnston and Kummu (2012) identified the following fundamental components of HEMs: flow volume, water level, connectivity, and flood dynamics for hydrology; habitat, wetland functions, fish, aquatic fauna and flora for ecology; resources from fish and aquatic products for livelihood; and costs and benefits of water uses for economic valuation The factors of space, time, management decision, and performance indicators are also needed as inputs for HEM (Heinz et al., 2007)

In terms of model design, hydro-economic models may be classified as holistic, modular, or Computable General Equilibrium (CGE) (Brouwer and Hofkes, 2008) The holistic approach is when hydrologic and economic components are embedded in a model (Cai et al., 2003) The holistic is used for determining interrelationships between

components to seek solutions to meet research objectives (Booker et al., 2012) An example of a holistic approach is the research of Booker and Young (1994) that

combined hydrologic data (inflow, evaporation, diversions within basins, and salinity) and economic data (demands of irrigation, municipal, and thermal energy) into a model

to simulate alternatives of water resource uses in Colorado River The modular

approach refers to using separate economic models and hydrologic models to achieve research objectives (Booker et al., 2012) Volk et al (2008) used the modular approach

to simulate ecological and economic models separately and then linked them together for obtaining research results Unlike the holistic and modular, the CGE approach

simulates economic models first and then links economic factors to hydrologic systems (Brouwer and Hofkes, 2008) CGE has been applied to evaluate new taxes on water demand for forestry and irrigation in South Africa that combine water flows and tax functions (van Heerden et al., 2008) and to simulate agricultural water uses in the San Joaquin Valley of California (Berck et al., 1990)

Hydro-economic models are often categorized by their application and function They may be particularly classified as simulation and optimization Simulation models use algorithms to simulate different scenarios The simulation was widely applied for diverse research, e.g., an assessment of the European Water Framework to European catchment (Thornes and Rowntree, 2006); research of Hydrogeological Unit ‘‘Eastern Mancha” in Spain (Martín de Santa Olalla et al., 2005); and hydrologic responses to scenarios considering groundwater and river flow (Lanini et al., 2004) Unlike the

simulation, optimization models were used for obtaining defined modeling objectives (Harou et al., 2009); where the modeling objective function includes benefits, costs, and

hydrologic factors (Booker et al., 2012) Examples of applying the optimization-type models include: simulating a watershed of the Olifants river in South Africa to optimize water delivery over time (Mukherjee, 1996); combining deterministic and stochastic approaches to optimize water allocation for irrigation (Paudyal and Manguerra, 1990); and dynamic programming to examine optimal land use (Dudley, 1988)

Integrated hydro-economic modeling have limitations (McKinney et al., 1999) First, the hydrologic uses simulation while the economic applies optimization Second, hydrologic and economic models do not have the same boundaries Third, the two modeling types have different time scales (i.e., hydrologic models uses days, months, or seasons while the economic models are years or longer) Fourth, hydro-economic models often have distinct model patterns that limit their application flexibility, e.g., energy (Odum, 1983), ecosystem (Jørgensen, 1992), water operation (Cohen and Sherkat, 1987), flood (Jonkman et al., 2008; Kazama et al., 2009; Verkade and Werner, 2011), and combined climate change, economic, and water resources (Jeuland, 2009) For example, the CBA evaluates water infrastructure investment while hydro-economic models consider operation and design of systems (Harou et al., 2009)

Limitations and distinct properties of HEMs require model assessment before use Several HEMs have been developed (Table 1-7) WEAP-based, CalSim II, AQUATOOL, WRAP, WBalMo, and GAMS-based are available hydro-economic models with

maintained and supported code These models include economic factors for simulation and may be appropriate for wetland research However, they are designed for large-scale applications and have been primarily used in the US or Western countries Thus,

they may not applicable in developing countries with limited hydrologic data

Table 1-7 Descriptions of some hydro-economic models

Models WEAP-based CalSim II AQUATOOL

-based

RiM-FIM Guan and

Hubacek (2008)

WRAP WBalMo

GAMS-based, CONOPT2 Developer SEI CDWR AIRH-

conservation

Operational criteria simulation

Assesses water scarcity

CEA for ecosystems management

Consumptions for water qualities

Assesses water availability

Assesses water scenarios

Assess water allocation strategies

Availability Yes Yes Yes Propriety Propriety Yes Yes Yes

Input data Land use,

climate, demand, crop

Priorities for water operation

Scarcity, pumping, operation

Geographical, infrastructure, investment costs, habitat

Water balance, water consumption

Water balance

& operating rules

Water balance and water demands

Irrigation, fish, wetland, industrial, hydropower Output

Assesses costs of scarcity

Determines cost- effectiveness for managing ecosystems

Assesses consumption

of water

Simulates water scenarios

Compares natural water yield

&

demands

Assesses wetland- related scenarios

Methods Application of

GIS and dynamics

Linear programing

Stochastic method

based and DEM

Satellite- hydrologic balance

Mass-Numerical methods

Carlo

Monte-The TCM and CVM

Location Middle

Guadiana basin, Spain

California State

Campo de Dalias, Spain

River Murray

in South Australia

North China Various Various Mekong

CDWR (2012)

Brouwer et

al (2009);

IIAMA-UPV (2012)

AIRH-Bryan et al

(2010)

Guan and Hubacek (2008)

Koch and Grünewald (2009); Wurbs and Dunn (1996)

IWPSR (2005)

Ringler and Cai (2006)

Notes: SEI is Stockholm Environment Institute-US Center, CDWR is California Department of Water Resources, WRPSR is Water Resources Planning & Systems Research, N/A is not available, WEAP-based is The Water Evaluation and Planning, CalSim II is The Water Resource Integrated Modeling System, AQUATOOL is Analysis and modeling of water resources systems management, RiM-FIM is The River Murray Floodplain Inundation Model, WRAP is Water Rights Analysis Package, WBalMo is Interactive Simulation System for Planning and Management

in River Basins, IWPSR is Institute for Water Resources Planning and Systems Research, and Availabilty is software availability