CN105260514A - Strong quantitative evaluation method for underground water pollution source - Google Patents

Abstract

The invention discloses a strong quantitative evaluation method for an underground water pollution source. The method comprises the following steps: collecting site information, simulating the concentration of a pollutant in the pollution source when reaching an underground water interface by HYDRUS-1D software, and calculating a reduction coefficient; carrying out strong master control factor screening of the underground water pollution source of the site by a grey correlation analysis method; building a multivariate linear regression equation through the screened strong master control factors of the underground water pollution source, and carrying out model checking and error evaluation; and carrying out a strong quantitative evaluation on the underground water pollution source of the site with limited information by the built multivariate linear regression equation. According to the method disclosed by the invention, a strong quantitative evaluation task of the underground water pollution source of the site with limited information can be effectively solved; a technical reference is provided for investigation and restoration of a pollution site; and the strong quantitative evaluation method is simpler than an HYDRUS-1D software simulation in input conditions, and is similar to an HYDRUS-1D software simulation result in output.

Description

Groundwater contamination source strength method for quantitatively evaluating

Technical field

The present invention relates to water environment protection field, relate more specifically to a kind of groundwater contamination source strength method for quantitatively evaluating, effectively to solve the groundwater contamination source strength quantitative evaluation task of data limited space, for the investigation of contaminated site and reparation provide Technical Reference.

Background technology

When oozing under the pollutant of earth's surface and pass through aeration zone, the physical chemistry biological respinses such as a series of dissolving, absorption, parsing can be there is, make pollutant arrive underground reservoir time structure, component, concentration and under ooze at the beginning of difference larger.According to country's ringborder management relevant regulations, groundwater contamination concentration value be pass judgment on whether groundwater quality exceeds standard, the degree that exceeds standard and formulate the key index of prophylactico-therapeutic measures.Science obtains Site characteristic pollutant levels exactly, has important guiding meaning for work such as Typical Sites groundwater environment effect appraise, site remediations.

In the report of relevant groundwater pollution simulation prediction both at home and abroad at present, rarely have and soil, aeration zone and underground water integrally carried out evaluation, be carry out simulation and forecast and evaluation analysis for the underground water with free-water level mostly, this have ignored the impact that aeration zone transforms contaminant transportation to a certain extent.But different contaminated sites has complicated pollution source of groundwater, and whether pollution source can have influence on groundwater quality, depend on different pollutant enter aeration zone after transportion and transformation, and pollutant levels value when finally entering underground reservoir.Therefore, soil, aeration zone and underground water are integrally carried out source strength evaluation to be necessary.According to countrydefinition in " environomental pollution source class code " (GB/T16706-1996), environomental pollution source refers to the occurring source that may cause environmental pollution.So, pollution source of groundwater just can be expressed as the occurring source that groundwater environment may be caused to pollute, and groundwater contamination source strength is exactly the influence degree of characteristic contamination to groundwater quality.

The domestic and international research for the evaluation of groundwater contamination source strength is relatively less at present, mostly Groundwater Pollution Evaluation is Groundwater Contamination Risk evaluation based on DRASTIC method, overlapped index method etc. and Evaluation of vulnerability, how not to consider the characteristic of pollution source and pollutant itself.Concentration value when Li Wei etc. utilize the different pollutant of HYDRUS-1D software simulation to arrive shallow aquifer under different aeration zone conditions, by the some analog result of HYDRUS-1D and MapGIS connected applications in large area, zonal pollution evaluation.But the setting of HYDRUS-1D analog parameter is comparatively loaded down with trivial details, operability higher for site investigation data requirement is general, is then difficult to realize for the incomplete place of data.

Summary of the invention

In view of this, the object of the present invention is to provide a kind of groundwater contamination source strength method for quantitatively evaluating.

For realizing this object, the invention provides a kind of groundwater contamination source strength method for quantitatively evaluating, comprising:

(1) collect place data, utilize HYDRUS-1D software to carry out reduction coefficient analog computation;

(2) utilize grey relational grade analysis method to carry out place groundwater contamination source strength main-control factors screening, filter out the factor higher with the reduction coefficient degree of association as independent variable, for the foundation of multiple linear regression equations;

(3) the groundwater contamination source strength main-control factors by filtering out sets up multiple linear regression equations, utilizes reduction coefficient compute dependent data to carry out model testing and error evaluation;

(4) multiple linear regression equations set up is utilized, the groundwater contamination source strength quantitative evaluation task of effective solution data limited space, for the investigation of contaminated site and reparation provide Technical Reference.

Known based on technique scheme, evaluation method of the present invention effectively can solve the groundwater contamination source strength quantitative evaluation task of data limited space, calculate groundwater contamination concentration value, pass judgment on whether groundwater quality exceeds standard, exceed standard degree, for the investigation of contaminated site and reparation provide Technical Reference.In addition, the present invention is in order to solve the constraint of place data for model prediction, propose a kind of groundwater contamination source strength method for quantitatively evaluating based on grey relational grade analysis, multiple linear regression model is set up based on grey relational grade analysis, to solve the groundwater contamination source strength quantitative evaluation that data information lacks place, comparatively HYDRUS-1D software simulation is more simple on initial conditions for the method, and output is similar to HYDRUS-1D software simulation result.

Accompanying drawing explanation

fig. 1for the signal of groundwater contamination source strength method for quantitatively evaluating figure.

Embodiment

For making the object, technical solutions and advantages of the present invention clearly understand, below in conjunction with specific embodiment, and reference accompanying drawing, the present invention is described in further detail.

The invention discloses a kind of groundwater contamination source strength method for quantitatively evaluating, comprising: first collect place data, calculate place reduction coefficient, according to grey relational grade analysis result determination main-control factors by HYDRUS-1D model; Then set up multiple linear regression equations for reduction coefficient, and carry out model testing and estimation of error, utilize regression equation new place reduction coefficient, finally calculate groundwater contamination concentration value, pass judgment on whether groundwater quality exceeds standard, exceed standard degree.The present invention effectively can solve the groundwater contamination source strength quantitative evaluation task of data limited space, for the investigation of contaminated site and reparation provide Technical Reference.

Particularly, groundwater contamination source strength method for quantitatively evaluating of the present invention, comprises the following steps:

(1) place Data Collection and reduction coefficient calculate

Place Data Collection comprises place aeration zone correlation parameter (soil types, soil thickness, unsaturated zone depth, terrain slope, infiltration coefficient, the soil weight etc.) needed for HYDRUS-1D software and contaminated sites thing correlation parameter (in longitudinal gas flow, free water coefficient of diffusion, adsorption coefficient etc.).HYDRUS-1D model is created on the basis of the models such as SUMATRA, WORM and SWMI by mechanisms such as United States Department of Agriculture, the saline and alkaline laboratories of the U.S. to develop, and this software can simulate the migration in saturated and unsaturated media such as One-Dimensional Water stream and solute and reaction in macroscopic view and micro-scale.Utilize reduction coefficient concentrated expression pollution source characteristic sum unsaturated soil structure to pollutant diafiltration to underground water mitigation, quantize aeration zone antifouling property.Calculate place groundwater contamination concentration value by HYDRUS-1D software, calculate reduction coefficient:

R _i＝C _i/C ₀

In formula: C ₀for the initial concentration of characteristic contamination, mg/L; C _ifor concentration value during characteristic contamination arrival water-bearing zone, mg/L.Time initial concentration is certain, the value arriving the larger reduction coefficient of concentration value in water-bearing zone is larger, the DeGrain of reduction; Reduction coefficient is less, and the effect of reduction is more obvious.Show that aeration zone is to the complete reduction of pollutant when the concentration value arriving water-bearing zone is 0, reduction coefficient is 0 with entering water concentration.

(2) grey relational grade analysis

Grey relational grade analysis is a kind of objective, qualitative, quantitative systematic analytic method, object is exactly prevailing relationship in the system of seeking between each factor, finds out the key factor affecting desired value, thus grasp the principal character of things, promotion and guidance system develop rapidly and effectively.

1. parameter assignment

Generally change the hydrogeological parameter in each place, concrete generalization thinking is: infiltration coefficient adopts the equivalent coefficient of permeability of stratified rock, and concrete formula is:

$K_v=\frac{{\mathrm{\Σ}}_{i=1}^n{M}_i}{{\mathrm{\Σ}}_{i=1}^n\frac{M_i}{K_i}};$

In formula: K _vfor the equivalent coefficient of permeability perpendicular to bedding angle, M _ifor each soil thickness, K _ifor each soil layer vertical hydraulic conductivity.

The soil weight adopts arithmetic mean; Dispersion coefficient and adsorption coefficient, by getting the method for empirical value, carry out assignment at an interval, adopt the mode of getting the value of overlapping interval to carry out parameter generally change for dispersion coefficient.

2. grey relational grade analysis

If k (k=1,2,3 ..., n) be place numbered sequence; X _ik () is factor of influence X _iabout the observation data in a kth place, then obtain factor of influence X _ihorizontal sequence be { X _i(k) | k=1,2,3 ..., n}, (i=1,2,3 ..., m), m is factor of influence quantity.

I () asks initial value picture and the average picture of each sequence:

Initial value picture ${\stackrel{\‾}{x}}_i={\mathrm{\Σ}}_{k=1}^n{X}_i\left(k\right),\stackrel{\‾}{R}={\mathrm{\Σ}}_{k=1}^{n}R\left(k\right),$

Average picture $R^{,}\left(k\right)=\frac{R\left(k\right)}{\stackrel{\‾}{R}};$

K=1 in formula, 2,3 ..., n; I=1,2,3 ..., m.

(ii) difference sequence is asked:

Δ _i(k)＝|R’(k)-X‘ _i(k)|

Δ _i＝{Δ _i(k)|k＝1，2，3，......，n}，(i＝1，2，3，......，m)。

(iii) two-stage maximum difference and lowest difference is asked:

$M=\underset{i}{Max}\underset{k}{Max}{\mathrm{\Δ}}_i\left(k\right),m=\underset{i}{Min}\underset{k}{Min}{\mathrm{\Δ}}_i\left(k\right).$

(iv) correlation coefficient is asked:

${\mathrm{\ζ}}_{Ri}\left(k\right)=\frac{m+\∂M}{{\mathrm{\Δ}}_i\left(k\right)+\∂M};$

k＝1，2，3，......，n；i＝1，2，3，......，m；

In formula: for resolution ratio, its role is to improve the significance of difference between correlation coefficient, $\∂\∈(0,1),$ Generally get $\∂=\mathrm{0.5.}$

(v) compute associations degree:

$r_{Ri}=\frac{1}{n}{\mathrm{\Σ}}_{k=1}^n{\mathrm{\ζ}}_{Ri}\left(k\right),i=1,2,3,......,m.$

Using the foundation that complexity two aspects of calculation of relationship degree result and place data acquisition are screened as groundwater contamination source strength main-control factors.Groundwater contamination source strength main-control factors, refers to and affects by force larger key factor to pollution source of groundwater, preferentially chooses the high factor of the degree of association as main-control factors, considers place data acquisition complexity when the degree of association is close.

(3) foundation of multiple linear regression equations

Multiple linear regression equations whether there is certain a kind of statistical analysis method that is linear or nonlinear relationship between the multiple independent variable of research and a dependent variable, and fundamental purpose is to investigate quantitative relation between these independent variable and dependent variables, explains that independent variable is on the impact of dependent variable.

Multiple linear regression equations process of establishing is:

1. utilize grey relational grade analysis to filter out the p relevant to a reduction coefficient main-control factors, set up dependent variable (reduction coefficient) and independent variable (main-control factors X _i, i=1,2 ..., the multiple linear regression model equation p).

2. reduction coefficient Y and p main-control factors X _ibetween multiple linear regression equations relational expression be:

Y＝b ₀+b ₁X ₁+b ₂X ₂+...+b _pX _p

In formula, b ₀, b ₁, b ₂..., b _pbeing the undetermined coefficient of a constant and p main-control factors respectively, is all estimated value.By least square method, can in the hope of the coefficient estimated value b of multiple linear regression equations.

(4) model testing and error evaluation

1. the test of fitness of fot

$R^2=\frac{ESS}{TSS}=1-\frac{RSS}{TSS},0\≤R^2\≤1$

R2 is more close to 1, and the degree of fitting of multiple linear regression equations is higher;

Wherein, ESS, RSS are respectively residual sum of squares (RSS) regression sum of square, and both equal TSS at addition.

2. the F inspection of regression equation conspicuousness

$F=\frac{ESS/p}{RSS/(n-p-1)}~F(p,n-p-1)$

Wherein, p is main-control factors number, and n is sample size, F _α(p, n-p-1) is F during conspicuousness α in tablenumerical value.

If F>=F _α(p, n-p-1), regression equation is significantly set up; If F < is F _α(p, n-p-1), regression equation is not remarkable;

3. the t inspection of regression coefficient conspicuousness

$t=\frac{{\mathrm{\&beta;}}_i}{S\left({\mathrm{\&beta;}}_i\right)};$

Wherein, β _ifor average of samples, S (β _i) be population average, for t during conspicuousness α/2 in tablenumerical value.

If show X _iconspicuousness effect is had to Y; If show X _ito Y without conspicuousness effect;

4. error analysis

Calculate the variance S of reduction coefficient raw sample data ², standard deviation S.Result and error is calculated, error of calculation variance S through back substitution _e ², posteriority difference ratio C=S _e ²/ S ², as small error possibility P=P{|Y-Y ' | during < 0.6745 × S}=1>=0.95, the error amount of model is little, reaches requirement.

(5) gained equation of linear regression is applied to new place, tries to achieve reduction coefficient, finally calculate groundwater contamination concentration value, pass judgment on whether groundwater quality exceeds standard, exceed standard degree.

Embodiment

(1) place Data Collection and reduction coefficient calculate

1. Data Collection

Collect the related data of 7 industries, 10 Typical Sites altogether, be respectively: Beijing landfill yard, xinjiangcertain power plant, Wujiang glass factory, Hunan scientific & technical corporation, Xiamen mining site, Henan open coal mine, Shanxi coal industry company, Shandong paper plant, the useless disposal site of Liaoning danger and the useless disposal site of Changsha danger.Hereafter use numbering 1,2,3 ..., 10 refer to 10 places.Pollution source and place aeration zone correlation parameter are shown in table 1with table 2.

table 1place aeration zone correlation parameter

table 2contaminated sites thing correlation parameter

2. reduction coefficient calculates

According to place data from investigation, use HYDRUS-1D software building model to calculate, simulation ammonia nitrogen enters the Transport And Transformation process after aeration zone, obtains the concentration value C that these place ammonia nitrogens arrive shallow aquifer _iafter, by C ₀with C _isubstitute into formula, obtain the reduction coefficient R of each place ammonia nitrogen _i, concrete numerical value is shown in table 3.

table 3the ammonia nitrogen reduction coefficient in ten places

Note: be the place of 0 for aeration zone reduction coefficient, coming from ammonia nitrogen, to enter water concentration be 0.

(2) grey relational grade analysis

1. parameter assignment

Choose initial ammonia nitrogen concentration of emission, unsaturated zone depth, terrain slope, infiltration coefficient, the soil weight, longitudinal gas flow, adsorption coefficient 7 factors as analysis indexes, use symbol X respectively ₁, X ₂, X ₃..., X ₇represent.The hydrogeological parameter in each place is generally changed, specifically sees table 4.

table 4the aeration zone reduction coefficient in ten places and each main affecting factors

Note: place numbering according to table 1in cited numbering

2. grey relational grade analysis

According to upper tablethe data message that lattice provide, utilizes Gray Correlation analysis and the higher main-control factors of the reduction coefficient degree of association.Calculate each factor of influence X _iwith the degree of association r of reduction coefficient R.The degree of association the results are shown in table 5.

table 5each Correlative Influence Factors degree of association result

Relatively learn: X ₄> X ₁> X ₂> X ₆> X ₅> X ₃> X ₇.

Using the foundation that complexity two aspects of calculation of relationship degree result and place data acquisition are chosen as Parameters in Regression Model.Adopt infiltration coefficient, ammonia nitrogen concentration of emission, unsaturated zone depth, the soil weight as 4 factors setting up multiple linear regression model.

(3) multiple linear regression model sets up result

1. the foundation of multiple linear regression equations

4 main-control factors are updated in following formula and obtain:

Y＝b ₀+b ₁X ₁+b ₂X ₂+b ₃X ₃+b ₄X ₄

Utilize SPSS software to calculate equation relational expression to be:

Y＝0.03-7.111×10 ^-7×X ₁-4.99×10 ^-6×X ₂+1.896×10 ^-5×X ₄

-0.019×X ₅

2. model testing interpretation of result

After SPSS software establishes regression equation in automated provisioning model testing parameter, test of fitness of fot parameter R ²be 0.986, quite close to 1, illustrate that equation model degree is very high.The significance test statistic F value of equation is 85.666.In this regression equation, independent variable totally 4, i.e. p=4; Totally 10 sampling of datas, i.e. n=10.Through tabling look-up, when conspicuousness α=0.05, F (4,5)=5.19; During α=0.01, F (4,5)=11.39; During α=0.005, F (4,5)=15.56.85.666 > > 15.56, therefore equation is very significant.

The inspection of parametric t value can be reduced to probability inspection, if the probability of t value is less than 0.05 just can determine that this independent variable is significant, SPSS provides data specifically to see table 6.

table 6coefficient gathers

After model testing, the regression equation relational expression finally obtained is:

Y＝0.03-7.111×10 ^-7×X ₁-4.99×10 ^-6×X ₂+1.896×10 ^-5×X ₄

-0.019×X ₅

3. error result analysis

The variance S of aeration zone reduction coefficient raw sample data ²=1.85 × 10 ^-4, standard deviation S=1.36 × 10 ^-2.After back substitution computing, back substitution result and error are shown in table 7, can be calculated error variance S _e ²=1.08 × 10 ^-6, posteriority difference ratio C=S _e ²/ S ²=0.0046 < 0.05, small error possibility P=P{|Y-Y ' | < 0.6745 × S}=1>=0.95, the error amount of this model is quite little.

table 7back substitution result

Note: place numbering according to table 1in cited numbering

(4) place application

With in state northcertain pig farm of side is example, verifies the multiple linear regression model of reduction coefficient.Pig farm is positioned at area, North China, and North China is commonly Quaternary Stratigraphic.Within the scope of the actual prospecting depth in factory site, stratum can be divided into artificial element soil, sand, gravel, sandy clay and fine sand from top to bottom, and distribution is comparatively continuous, stable.This district's underground water is mainly diving, current plant area bury of groundwater about 60 meters.According to the Surveillance in pig farm in May, 2014, in plant area's waste water, the initial ejection concentration of ammonia nitrogen is 261mg/L, and through the sampling and testing of groundwater monitoring well, in shallow ground water, ammonia nitrogen concentration monitor value is less than 2.0 × 10 ^-3mg/L.

1. the calculating of parameter assignment and reduction coefficient

Parameters in Mathematical Model is arranged to be seen table 8with table 9.

table 8different soils layer correlation parameter

table 9different soils layer Nitrogen transformation analog parameter

After the simulation of HYDRUS-1D software, result of calculation is: 9.142 × 10 ^-4mg/L.Obtaining reduction coefficient according to reduction coefficient formulae discovery is 3.5 × 10 ^-6.

The numerical value providing four parameters in conjunction with pig farm actual conditions is respectively: X ₁for 261mg/L, X ₂for 6000cm, X ₄for 300cm/d, X ₅for 1.4g/cm ³, the value that the multiple linear regression equations set up calculates Y is-2.1 × 10 ^-2.Can think ammonia nitrogen arrive water-bearing zone time concentration be 0.

2. evaluation result contrast

According to the Surveillance in pig farm in May, 2014, in shallow ground water, ammonia nitrogen concentration monitor value is 1.8 × 10 ^-3mg/L, the HYDRUS-1D ammonia nitrogen concentration analogue value is 9.1 × 10 ^-4mg/L, and during ammonia nitrogen arrival shallow aquifer, concentration returns value is 0.I class Water quality ammonia nitrogen concentration value in standards of water quality (GB/T14848-93) under base area, it is up to standard that the ammonia nitrogen concentration monitor value in this place, the analogue value and regressand value all show ammonia nitrogen concentration, and three's source strength evaluation conclusion conforms to.

Above-described specific embodiment; object of the present invention, technical scheme and beneficial effect are further described; be understood that; the foregoing is only specific embodiments of the invention; be not limited to the present invention; within the spirit and principles in the present invention all, any amendment made, equivalent replacement, improvement etc., all should be included within protection scope of the present invention.

Claims (6)

1. a groundwater contamination source strength method for quantitatively evaluating, comprises following each step:

Step 1: the data gathering every profession and trade Typical Sites, to utilize in HYDRUS-1D software simulation pollution source under pollutant arrival point concentration during water termination, calculates reduction coefficient;

Step 2: utilize grey relational grade analysis method to carry out place groundwater contamination source strength main-control factors screening, filter out the factor high with the described reduction coefficient degree of association as independent variable;

Step 3: set up multiple linear regression equations by the described groundwater contamination source strength main-control factors filtered out;

Step 4: utilize the described multiple linear regression equations set up, the quantitative evaluation of groundwater contamination source strength is carried out to target place.

2. method for quantitatively evaluating as claimed in claim 1, wherein, the reduction coefficient R described in step 1 _icomputing formula is as follows:

R _i＝C _i/C ₀

In formula: C ₀for the initial concentration of characteristic contamination, mg/L; C _ifor concentration value during characteristic contamination arrival water-bearing zone, mg/L.

3. method for quantitatively evaluating as claimed in claim 1, grey relational grade analysis method is wherein utilized to carry out in the step of place groundwater contamination source strength main-control factors screening described in step 2, comprise the step of the hydrogeological parameter in each place being carried out to generalization: infiltration coefficient adopts the equivalent coefficient of permeability of stratified rock, and concrete formula is:

$K_v=\frac{{\mathrm{\&Sigma;}}_{i=1}^n{M}_i}{{\mathrm{\&Sigma;}}_{i=1}^n\frac{M_i}{K_i}};$

In formula: K _vfor the equivalent coefficient of permeability perpendicular to bedding angle, M _ifor each soil thickness, K _ifor each soil layer vertical hydraulic conductivity.

4. method for quantitatively evaluating as claimed in claim 1, the step wherein utilizing grey relational grade analysis method to carry out place groundwater contamination source strength main-control factors screening described in step 2 comprises:

If k (k=1,2,3 ..., n) be place numbered sequence; X _ik () is factor of influence X _iabout the observation data in a kth place, then obtain factor of influence X _ihorizontal sequence be { X _i(k) | k=1,2,3 ..., n}, (i=1,2,3 ..., m), m is factor of influence quantity;

1. initial value picture or the average picture of each sequence is asked:

K=1 in formula, 2,3 ..., n; I=1,2,3 ..., m;

2. difference sequence is asked:

Δ _i(k)＝|R’(k)-X‘ _i(k)|

Δ _i＝{Δ _i(k)|k＝1，2，3，......，n}，(i＝1，2，3，......，m)；

3. two-stage maximum difference and lowest difference is asked:

$M=\underset{i}{Max}\underset{k}{Max}{\mathrm{\&Delta;}}_i\left(k\right),m=\underset{i}{Min}\underset{k}{Min}{\mathrm{\&Delta;}}_i\left(k\right);$

4. correlation coefficient is asked:

${\mathrm{\&zeta;}}_{Ri}\left(k\right)=\frac{m+\&part;M}{{\mathrm{\&Delta;}}_i\left(k\right)+\&part;M}$

k＝1，2，3，......，n；i＝1，2，3，......，m；

In formula: for resolution ratio, its role is to improve the significance of difference between correlation coefficient, generally get

5. compute associations degree:

$r_{Ri}=\frac{1}{n}{\mathrm{\&Sigma;}}_{k=1}^n{\mathrm{\&zeta;}}_{Ri}\left(k\right),i=1,2,3,......,m.$

5. method for quantitatively evaluating as claimed in claim 1, wherein, the step setting up multiple linear regression equations by the described groundwater contamination source strength main-control factors filtered out described in step 3 comprises:

1. utilize grey relational grade analysis to filter out the p relevant to a reduction coefficient main-control factors, set up dependent variable and the main-control factors X of reduction coefficient _i, i=1,2 ..., the multiple linear regression model equation between independent variable p);

2. reduction coefficient Y and p main-control factors X _ibetween multiple linear regression equations relational expression be:

Y＝b ₀+b ₁X ₁+b ₂X ₂+…+b _pX _p

In formula, b ₀, b ₁, b ₂..., b _pbeing the undetermined coefficient of a constant and p main-control factors respectively, is all estimated value; By least square method, try to achieve the coefficient estimated value b of described multiple linear regression equations.

6. method for quantitatively evaluating as claimed in claim 1, wherein in step 3 after the described described groundwater contamination source strength main-control factors by filtering out sets up the step of multiple linear regression equations, also comprise the step utilizing the data of groundwater contamination source strength main-control factors to carry out model testing and error evaluation.