an analysis of covariance to determine the effect of tplms...
TRANSCRIPT
An Analysis of covariance to determine the effect of TPLMS
in Masan Bay, Korea
Yoon Ju Yi*, Jungho Nam,
Won Keun Chang
Korea Maritime Institute
28th August 2011 / Baltimore, USA
Contents
The features of Masan Bay1
Total Pollution Load Management System(TPLMS) in Masan Bay2
Test for effect of TPLMS using ANCOVA3
Conclusion & Future study4
2
Length: ca. 10km, Width of mouth : ca. 5.0km Average depth: 10~15m
Masan Bay
(2011)
Masan Bay
(1899)
Free trade Zone
New town
Thermoelectric
Power Plant
Port & Industrial
Facilities
Wharf
& Port
New town
New town
& Port Facilities
Masan Bay SMA
(MLTM, 1998)4
Watershed : 263.98 sq. kilometersWatershed population : ca. 1 M people Population density : 3,167 individuals/km2 (nation : 460)
Geographical Features of Masan Bay
History of Environmental management of Masan Bay
Port of Masan : 0.4 km2 (~’50s)
(‘75) Gapobeach closure
Han-Il Fiber : 0.23 km2 (~’66)Han-KukSteel : 0.17 km2 (~’67)
Free Trade Zone(~’70)
(‘79) Public Ban for gathering Clam and Mussel
Massive HAB occurrence(‘81) (‘82) SMA designatedChanwonInd. Complex(‘80)
(’90~) Polluted sediment dredged(’93) Sewage treatment plant constructed
(’00) SMA re designated including watershedDaeHanTransportation : 0.15 km2 (‘02)
West Port Zone : 1.36 km2 (‘11)
(’04) SMA management plan published
(’07) STP facilities improved
(’02) Jin haeSWP constructed
(’07~) TMDL Implemented
1950
1960
1970
1980
1990
2000
2010
5
Dissolved Oxygen Conc. (Bottom Layer)
Occurrence of Hypoxia (below 2㎎/L) amounts to 27% of the national total.
1997 1998 1999 2000 2001 2002 2003 2004 2005 Total
National 9 2 2 4 - 3 4 6 7 37
Masan 1 2 1 2 - 2 - - 3 11
% 11.1 100.0 50.0 50.0 - 66.7 - - 28.6 27.0
Red Tide (or HAB) Occurrence
Red Tides in Masan Bay occur 52 times, amounting to 14.2% of the national total (’00~’05).
The ‘80-’81 massive Red-Tide was reported in Masan Bay(the 1st case in Korea)
Total 2000 2001 2002 2003 2004 2005
Red Tide
(Occurrence)
Masan 51 9 6 10 8 9 9
National 358 85 59 59 48 69 38
% 14.2 10.6 10.2 16.9 16.7 13.0 23.7
Chemical Oxygen Demand (COD)
NFRDI reported the water quality of Masan Bay : 10-year Median of COD conc. in Masan Bay, 2.4㎎/L, is twice higher than that of the nation, 1.15 ㎎/L. (3rd Grade)
Median(year) Feb. May Aug. Nov.
COD(㎎/L)
Masan Bay 2.40 2.36 2.54 2.65 2.08
National 1.15 1.12 1.13 1.27 1.06
Status of Marine Environment of Masan Bay
6
Land Use Planning without considering the Carrying Capacity of Masan Bay- 42 coastal development projects still planned
- Conflict between Masan city, MLTM, and local NGOs, on the selection and management of a
dumping site for dredged material
Weak integrated management system Lack of comprehensive approach to various pollution sources and input pathways
Lack of coordination between economic vision and environment management goals
Conflict between ministries on jurisdiction lack of cooperation
Lack of systematic strategic action plan Implementation of environment management measures without priority, shared vision and goals, and
social consensus
ex) Limited effect of dredging without effective control of land-based pollution sources
Less integrated and partnership-based management entity Coastal Environment Management Plan of the Masan-bay SMA established through cooperation
among related ministries and local governments
But, no integrated management entity with participatory planning and implementation mechanisms
yet
Analysis of Constrains of TPLMS Introduction
Principles and Strategies for establishing TPLMS
Marine Ecosystem Model
Determination of Loads of Municipalities
Estimation of Max. Allowable Loads
Estimation of LoadsMonitoring Program
14 rivers + 2 swage plants
Determination of Pollution Load Reduction Target
Determination of W/Q target (2.4 or 2.5)
Verification of MoS, and other parameters
Determination of Allocation Principles
Development of New Reduction Plan
Assessment
1st TPLMS Program
Implementation Plan
Adaptation of Implementation Plans
COD(N, P)
Firm Scientific Platform
Source : Chang et al.,(2008)
Principles and Strategies for establishing TPLMS
9
Proposed Investments for Loads Reduction Prgms(as of Oct. 2008)
Unit : Kg/DayExpected annual Load Reduction (Target : 4,245 kg/day) (as of Oct. 2008)
0%
75%
99%106%
58%
93%
Programs ‘07 ‘08 ‘09 ‘10 ‘11 Sum
1. Sewage Coverage Enhancement 285 285
2. River Management 9.63 29.3 181.96 25.07 245.96
3. Improvement of Sewage Treatment
Facility and Capacity92.55 676 768.55
4. Effluents Management 738 738
5. Cancellation of Development plans 2,456 2,456
Total (cumulative) 3,203.63 3,203.63 3232.93 3,507.44 4,493.51 4,493.51
Loads Reduction Prgms. ‘07 ‘08 ‘09 ‘10 ‘11 Sum
1. Sewage Coverage Enhancement 3,000 3,066 3,066 9,132
2. River Management 8,224 14,804 25,310 20,868 1,200 70,406
3. Improvement of Sewage Treatment
Facility and Capacity
1,054 2,100 9,200 11,900 2,100 26,354
4. Effluents Management 0 0 0 0 0 0
5. Cancellation of Development plans 2,100 0 0 0 0 0
Total 11,378 16,904 37,510 35,834 6,366 107,992
Unit : M KRW
0%
2%
92%
100%
26%
Pollution Load Reduction Programs and their Costs of TPLMS
Classification# of
Samplingitems
Details Sampling peorid# of
Sampling stations
Coastal Environ-
ment
Water Quality 14
Water temperature, Salinity, Dissolved Oxygen(DO), Hydrogen ion concentration(pH), Chemical Oxygen Demand(CODMn), SuspendedSolids(SPM), Dissolved Inorganic Nitrogen(DIN), Dissolved Inorganic Phosphorous(DIP), Total Nigrogen(TN), Total Phosphorous(TP), Silicate-silicon(SiO2), Phytoplankton Biomass(Chl-a), Total Organic Carbon(TOC), Particulate Organic Carbon(POC) etc.
1 time per month(variability)
13
Sediments 8Ignition Loss(IL), Sediment COD, Acid Volatile Sulfide(AVS), Organicmatter(TOC, TN, TP), Plant pigment(Chl-a), grain size composition of sediment
1 time per month(variability)
13
Ecosystem 3The number of appearance species, Species composition, Population density, Diversity Index (Targets : Zoo∙Phytoplankton, Zoobenthos)
Seasonal 13
Features of Seawater flow
2 Tidal current and Water-exchange ratio Seasonal 2
Releasing /Settling rate of
Sediment2/4
Releasing rate : 1) Oxygen Requirement(SOD)2) Releasing rate of nutrient(DIN&DIP flux)Settling rate : 1) Total settling rate2) Ratio of Carbon, Nitrogen and Phosphorous in settling matters3) Settling velocity of particulate carbon, nitrogen and phosphorous
Seasonal 4
River/ Basin
Environ-ment
River 20Precipitation, Flux, Water temperature, Hydrogen ion concentration(pH), Salinity, Electrical conductivity, Dissolved Oxygen(DO), SuspendedSolids(SS), Biochemical Oxygen Demand (BOD5), Chemical Oxygen Demand(CODMn), Total nitrogen(TN), Total phosphorous(TP), Dissolved Inorganic Nitrogen(Nitrite nitrogen, Ammonia-nitrogen, Nitrate-nitrogen), Dissolved Inorganic Phosphorous(DIP), Silicate-silicon(SiO2), Total Organic Carbon(TOC), Particulate Organic Carbon(POC), Dissolved Organic Carbon(DOC)
1 time per month(10-day intervals
In summer)17
Outfall discharged
from Treatment Plant
201 time per month(10-day intervals
In summer)2
12
Summary of Monitoring(Rivers and Bay water)
Methodology(ANCOVA) – 1/2
Analysis of Covariance(ANCOVA) A general linear model with a continuous outcome variable(quantitative, scaled) and two or more
predictor variables where at least one is continuous(quantitative, scaled) and at least one is categorical(nominal, non-scaled). ANCOVA is a merger of ANOVA and regression for continuous variables. ANCOVA test whether certain factors have an effect on the outcome variable after removing the variance for which quantitative predictors(covariates) account. The inclusion of covariates can increase statistical power because is accounts for some of the variability.
Assumptions Like any statistical procedure, the interpretation of ANCOVA depends on certain assumptions about the
data entered into the model. For instance, the F-test assumes that the errors are normally distributed and homoscedastic. Since ANCOVA is a method based on linear regression, the relationship of the dependent variable to the independent variable(s) must be linear in the parameters.
Simplifying assumption(not necessary to run ANCOVA): homogeneity of regression which says that the relationship between the covariate and the dependent variable should be similar across all groups of the independent variable.
Power consideration While the inclusion of a covariate into an ANOVA generally increases statistical power by accounting for
some for the variance in the dependent variable and thus increasing the ratio of variance explained by the independent variables, adding a covariate into ANOVA also reduces the degrees of freedom. Accordingly, adding a covariate which accounts for very little variance in the dependent variable might actually reduce power.
(Source : http://en.wikipedia.org/wiki/Analysis_of_covariance)14
15
Methodology(ANCOVA) – 2/2
Statistical test of whether or not the means of several groups are all equal
ANOVA
A statistical technique for estimating the relationship among variables
Regression
Statistical Test of whether or not the means of several groups are equal controlling potential quantitative variable
ANCOVA
To compare the relationship between a dependent(Y) and continuous independent variable(X1) for different levels of one or more categorical variables(X2) : Y = μ + α(X1) + β(X2) + γ (X1*X2) + ε
When we do use ANCOVA?
Outline of the ANCOVA test
Model formula COD Concentration(Surface/Bottom) = α + β1(Year) + β2(COD Load)
+ γ(Year*COD Load) + ε, where ε~iid(0,σ2).
Variables Concentration(Y) : COD (TN, TP) Concentration of Masan bay water
COD Load(X1) : COD (TN, TP) Load of Rivers surrounding Masan bay
Year(X2) : A categorical variable classifying the period implementing TPLMS or not
• 0 : Before implementing TPLMS (2005-2006)
• 1 : After implementing TPLMS (2010)
Significance Level : 0.05
16
Procedure of analysis
17
Full model: Y X Xij i i ij i ij( )
0 : constantiH
ANCOVA model:Y X Xij i ij i ij( )
0:
, allfor 0:
02
01
i
i
H
iH
Separate Regression
Multiple ComparisonCommon Regression: Y X Xij ij ij( )
01
02
: 0,
: 0.
H
H
Accepted 0H
Accepted 01H
Rejected 0H
Rejected 01H
(Modified from Lecture materials of Ottawa University 2001)
X1
Y
X1
Y
Test for differences among slopes
Test for differences among levels
Test for significance of parameters
About COD Bottom Surface
category
model 24.52** 11.58**
Covariate 0.01 0.04
Year 51.54** 26.85**
slope 8.61** 2.47
Estimating slope difference between two periods 2.09** 2.90**
18
Result of the ANCOVA test
(Statistically significant at *p<0.05 or ** p<0.01)
• Main results of ANCOVA test1) Base material : COD Find out statistically significant
differences(Surface : 2.90mg/L, Bottom : 2.09mg/L)2) Base material : TN, TP No significant results
• Outcomes of TPLMS(locally)1) Reduction of pollutant load from
rivers (COD standard. 2000kg/day)2) Restoration of ecosystem :
Re-appearance of natural monuments such as mandarin duck, sesarma intermedium (crab), etc.
3) Not red-tide found during the past 3 years
• Outcomes of TPLMS(nationally) The 1st domestic success case
applying TPLMS• Outcomes of TPLMS(nationally) One of the best examples
implementing TPLMS among Asia
Conclusion Future study
• The 1st TPLMS : Focused on Reduction of COD load ⇒ prioritize in controlling point sources
• Planning treatment measure for TN, TP when establishing 2nd TPLMS in 2012 ⇒ Tighten the policy for controlling non-point sources
• Expanding the study: Establishing plans and measures reflecting the scientific research results of Masan bay1) Planning for urban regeneration2) National management plan for
pollution from land-based source※ Two drafts will be published in 2013
• Expanded application of the TPLMS in other areas1) Shihwa starts in 20132) Busan starts in 2015
22
Harbor
CEMAs
Summary of Environmental
Characteristics of SMAs’
Watershed- Shihwa Lake-Incheon: a city,
industrial complex, harbor, dike
- Kwangyang : a city, industrial
complex, harbor, Sumjin River
- Busan : a city, industrial complex,
harbor, open sea
- Ulsan : a city, industrial complex,
harbor, open sea
- Masan : a city, industrial complex,
harbor, semi-closed bay
Environmental Pressures of SMA’s Watershed
24
26
’05~’10, COD Load of Rivers vs. COD concentration of Masan Bay
Reducing pollution load in river
Improving water quality in bay