Mixing regime as a key factor to determine DON formation in drinking water biological treatment

小知识点：1：水中杂质包括溶解物，胶体和悬浮物。

尺寸分别为0.1nm-1nm,10nm-100nm,1μm-1mm。

2：悬浮物尺寸较大，易于在水中下沉或上浮，胶体颗粒尺寸很小，在水中长期静置也难下沉。

水中所含胶体通常含有粘土，某些细菌和病毒，腐蚀质及蛋白质等。

悬浮物和胶体是使水产生浑浊现象的根源。

悬浮物和胶体是饮用水处理的主要去除对象。

溶解杂质包括有机物和无机物两类。

3：地下水硬度高于地表水。

我国地下水的总硬度通常在60-300mg/L(以Cao计)之间，少数地区可达到300-700mg/L。

4：生活饮用水卫生标准GB2006，一共106项指标（增加5项），包括感官性状和一般化学指标、毒理学指标、细菌学指标和放射性指标。

浑浊度不超过1度；游离余氯在与水接触30min后不应低于0.3mg/L，集中式给水除出水厂应符合上述要求外，管网末梢水不应低于0.05mg/L。

5：软化方法主要有:离子交换法和药剂软化法。

6：对于不受污染的地表水，饮用水处理对象主要是去除水中悬浮物、胶体和致病微生物。

常规处理十分有效（即混凝、沉淀、过滤、消毒）；对于污染水源，需在常规处理工艺前加预处理，后加深度处理。

预处理和深度处理的对象主要是水中有机污染物。

7：预处理方法主要有：粉末活性炭吸附法，臭氧或高锰酸钾氧化法，生物氧化法等；深度处理主要有:粒状活性炭吸附法，臭氧-粒状活性炭联用法或生物活性炭法，化学氧化法，光化学或超声波-紫外线联用法等物理化学氧化法，膜滤法等。

8：传递机理可分：主流传递，分子扩散传递和紊流扩散传递。

9：理想反应器模型包括：完全混合间歇式反应器（CMB型）；完全混合连续式反应器（CSTR型），推流式反应器（PF型）。

10：混凝过程涉及三方面的问题：水中胶体粒子的性质，混凝剂在水中的水解物种以及胶体粒子与混凝剂之间的相互作用。

11：胶体失去稳定性的过程叫称“凝聚”，脱稳胶体相互聚集称“絮凝”。

“混凝”是凝聚和絮凝的总称。

Abstract This paper describes the principles of a novel 3D PIV system based on the illumination,recording and reconstruction of tracer particles within a 3D measurement volume.The technique makes use of several simultaneous views of the illuminated par-ticles and their 3D reconstruction as a light intensity distribution by means of optical tomography.The technique is therefore referred to as tomographic particle image velocimetry (tomographic-PIV).The reconstruction is performed with the MART algo-rithm,yielding a 3D array of light intensity discretized over voxels.The reconstructed tomogram pair is then analyzed by means of 3D cross-correlation with an iterative multigrid volume deformation technique,returning the three-component velocity vector distri-bution over the measurement volume.The principles and details of the tomographic algorithm are discussed and a parametric study is carried out by means of a computer-simulated tomographic-PIV procedure.The study focuses on the accuracy of the light intensity field reconstruction process.The simulation also identifies the most important parameters governing the experi-mental method and the tomographic algorithm parameters,showing their effect on the reconstruction accuracy.A computer simulated experiment of a 3D particle motion field describing a vortex ring demon-strates the capability and potential of the proposed system with four cameras.The capability of the tech-nique in real experimental conditions is assessed with the measurement of the turbulent flow in the near wake of a circular cylinder at Reynolds 2,700.1IntroductionThe instantaneous measurement of the 3D velocity field is of great interest to fluid mechanic research as it enables to reveal the complete topology of unsteady coherent flow structures.Moreover,3D measurements are relevant for those situations where the flow does not exhibit specific symmetry planes or axes and sev-eral planar measurements are required for a sufficient characterization.In particular this applies to flow tur-bulence,which is intrinsically 3D and its full descrip-tion therefore requires the application of measurement techniques that are able to capture instantaneously its 3D structure,the complete stress tensor and the vor-ticity vector.The advent of PIV and its developments (stereo-PIV,Arroyo and Greated 1991;dual-planestereo-PIV,Ka¨hler and Kompenhans 2000)showed the capability of the PIV technique to quantitatively visu-alize complex flows.Several different methods were also proposed to achieve a 3D version of the technique(scanning light sheet,Bru¨cker 1995;holography,Hinsch 2002;3D PTV,Maas et al.1993).A novel system for 3D velocity measurements based on tomographic reconstruction of the 3D particle dis-tribution is investigated in the present study.Record-ings of particle images from an illuminated volume taken from several viewing directions simultaneouslyG.E.Elsinga (&)ÁF.Scarano ÁB.W.van Oudheusden Department of Aerospace Engineering,Delft University of Technology,Kluyverweg 1,Delft 2629HS,The Netherlands e-mail:g.e.elsinga@tudelft.nlB.WienekeLaVision GmbH,Go¨ttingen,Germany Exp Fluids (2006)41:933–947DOI 10.1007/s00348-006-0212-zRESEARCH ARTICLETomographic particle image velocimetryG.E.Elsinga ÆF.Scarano ÆB.Wieneke ÆB.W.van OudheusdenReceived:13April 2006/Revised:4September 2006/Accepted:19September 2006/Published online:11October 2006ÓSpringer-Verlag 2006are used to reconstruct the3D light intensity distribu-tion.The method is therefore referred to as tomo-graphic particle image velocimetry(tomographic-PIV). Provided that two subsequent exposures of the particle images are obtained,the measurement technique re-turns the instantaneous velocityfield within the mea-surement volume by means of3D particle pattern cross-correlation.The motivation for such a3D PIV system is pre-sented in the next section together with an overview of current alternative techniques.Then the working principle and the tomographic reconstruction method are discussed in detail,followed by a numerical simu-lation study assessing the effect of the relevant exper-imental and reconstruction parameters and the full-scale capabilities of tomographic-PIV.The study concludes with the application of the proposed tech-nique to a cylinder wakeflow in the turbulent regime. The measurement of3D coherent structures within the wake is discussed and assessed.2Current3D PIV techniquesAmong the different3D velocimetry techniques pres-ently available holographic-PIV has received most attention(Hinsch2002;Chan et al.2004).It uses the interference pattern of a reference light beam with light scattered by a particle,which is recorded on a holo-gram,to determine the particle location in depth.The in-plane position in principle is given by the position of the diffraction pattern in the image.Illumination of the hologram with the reference light beam reproduces the original light intensityfield in the measurement volume at the time of recording,the intensity being highest at the original particle location.The reconstructed inten-sityfield is scanned by a sensor,e.g.a CCD,to obtain a digital intensity map,which can be used for cross-cor-relation yielding the velocityfield.So far holographic-PIV has shown a great potential in terms of a high data yield.However,its drawbacks are that the recording medium is a holographicfilm requiring wet processing, which makes the process time consuming and somehow inaccurate due to misalignment and distortion when re-positioning the hologram for the object reconstruction. The technique was successfully applied to measure a vortex ring in air,the wake of a mixing tab in water(Pu and Meng2000),a cylinder wakeflow in air and a free air nozzleflow(Herrmann et al.2000)returning large numbers of vectors(up to92,000using individual par-ticle pairing,Pu and Meng2000).Instead of recording on a photographic plate,the hologram can also be captured directly by a CCD sensor(Digital-Holographic-PIV,Coe¨tmellec et al. 2001).In that case the light intensity distribution in the measurement volume is evaluated numerically,usually by solving the Fresnel diffraction formula on the hologram(near-field diffraction,Pan and Meng2002). CCD sensors,however,have a very limited spatial resolution compared to the photographic plate ret-urning about2to3orders less particle images and velocity vectors.Moreover,the large pixel pitch requires that the recording is obtained at a relatively small angle(a few degrees between reference beam and scattered light)in order to resolve the interference pattern,hence strongly limiting the numerical aperture and depth resolution(Hinsch2002).The scanning-PIV technique is directly derived from standard2C or stereo PIV with the light sheet scanning through the measurement volume(Bru¨cker1995).The volume is sliced by the laser sheet at sequential depth positions where the particle image pattern is recorded. The second recording at that depth position can be taken either directly after thefirst or after the complete scan of the volume.The procedure returns planar velocityfields obtained slightly shifted in space and time,which can be combined to return a3D velocity field.The strong point of this technique is the high in-plane spatial resolution and the fact that the analysis of the recordings is straightforward.However,scanning-PIV requires high-repetition systems(kHz)to ensure that the complete volume recording is almost simul-taneous.The underlying hypothesis of scanning-PIV is that the volume scanning time needs to be small if compared with the characteristic time scale of the investigatedflow structure.Cameras with high a recording rate are thus required,which is not a prob-lem in low speedflow as shown by Hori and Sakaki-bara(2004)measuring a turbulent jet in water. However,the technique is unsuited for airflows and in particular in high speedflows.Moreover,high repeti-tion rate lasers are expensive and provide relatively low pulse energy.It should also be remarked that the experimental setup is significantly more complicated by the addition of a scanning mechanism.As an alternative to scanning,dual plane stereo PIV (Ka¨hler and Kompenhans2000)records the particle images in different planes simultaneously using light polarization or different colors to distinguish the scat-tered light from the two planes.In principle,mea-surements can be performed over more than two planes with each plane requiring a double-pulse laser, however separation by polarization is the most com-monly adopted solution and is possible only over two planes.Furthermore,using different colors complicates the optical arrangement.Relatively recent options are3D PTV(Maas et al. 1993)and defocusing PIV(Pereira et al.2000).These techniques rely on the identification of individual par-ticles in the PIV recordings.The exact position of the particle within the volume is given by the intersection of the lines of sight corresponding to a particle image in the recordings from several viewing directions(typi-cally three or four).The implementation of the particle detection and location varies with the methods.In comparison with the previous two methods,the3D-PTV approach offers the advantage of being fully digital and fully3D without the requirement for mov-ing parts.The velocity distribution in the volume is obtained from either particle tracking or by3D-cross-correlation of the particles pattern(Schimpf et al. 2003).However,the procedure for individual particle identification and pairing can be complex and as it is common for planar PTV several algorithms have been proposed,which significantly differ due to the prob-lem-dependent implementation.The main limitation reported in literature is the relatively low seeding density to which these techniques apply in order to keep a low probability of false particle detection and of overlapping particle images.Moreover,the precision of the volume calibration or in the description of the imaging optics isfinite.This means the lines of sight for a particle almost never truly intersect and an inter-section criterion is needed.Consequently the maxi-mum seeding density in3D PTV is kept relatively low. Maas et al.(1993)report a seeding density of typically 0.005particles per pixel for a three camera system.The development of the proposed tomographic-PIV technique is motivated by the need to achieve a3D measurement system that combines the simple optical arrangement of the photogrammetric approach with a robust particle volume reconstruction procedure, which does not rely on particle identification.As a consequence the seeding density can be increased,with respect to3D PTV,to around0.05particles per pixel (as will be shown later),which is not far from the particle image density used in planar experiments.As a consequence theflowfield within the3D domain can be represented with a number of velocity vectors comparable or slightly higher than obtained in planar PIV.Furthermore,the proposed technique features the instantaneousflowfield measurement,as opposed to scanning PIV,opening the possibility to perform3D measurements in several conditions irrespective of the flow velocity.Finally the introduction of high-repeti-tion rate PIV hardware is expected to further extend the measurement technique capability to a3D time-resolvedflow diagnostic tool as already demonstrated in a recent study(Schro¨der et al.2006).3Working principle of tomographic-PIVThe working principle of tomographic-PIV is sche-matically represented in Fig.1.Tracer particles im-mersed in theflow are illuminated by a pulsed light source within a3D region of space.The scattered light pattern is recorded simultaneously from several view-ing directions using CCD cameras similar to stereo-PIV,applying the Scheimpflug condition between the image plane,lens plane and the mid-object-plane.The particles within the entire volume need to be imaged in focus,which is obtained by setting a proper f/#.The3D particle distribution(the object)is reconstructed as a 3D light intensity distribution from its projections on the CCD arrays.The reconstruction is an inverse problem and its solution is not straightforward since it is in general underdetermined:a single set of projections can result from many different3D objects.Determining the most likely3D distribution is the topic of tomog-raphy(Herman and Lent1976),which is addressedin Fig.1Principle of tomographic-PIVthe following section.The particle displacement(hence velocity)within a chosen interrogation volume is then obtained by the3D cross-correlation of the recon-structed particle distribution at the two exposures.The cross-correlation algorithm is based on the cross cor-relation analysis with the iterative multigrid window (volume)deformation technique(WIDIM,Scarano and Riethmuller2000)extended to3D.The relation between image(projection)coordi-nates and the physical space(the reconstruction vol-ume)is established by a calibration procedure common to stereo PIV.Each camera records images of a cali-bration target at several positions in depth throughout the volume.The calibration procedure returns the viewing directions andfield of view.The tomographic reconstruction relies on accurate triangulation of the views from the different cameras.The requirement for a correct reconstruction of a particle tracer from its images sets the accuracy for the calibration to a frac-tion of the particle image size(see Sect.5).Therefore, a technique for the a-posteriori correction for the sys-tem misalignment can significantly improve the accu-racy of reconstruction(Wieneke2005).The mapping from physical space to the image coordinate system can be performed by means of either camera pinhole model(Tsai1986)or by a third-order polynomial in x and y(Soloff et al.1997).4Tomographic reconstruction algorithmThe novel aspect introduced with tomographic-PIV is the reconstruction of the3D particle distribution by optical tomography.Therefore,a separate section is devoted to the tomographic reconstruction problem and algorithms for solving it.By considering the properties of the measurement system,it is possible to select a-priori the reconstruc-tion method expected to perform adequately for the given problem.First the particle distribution is dis-cretely sampled on pixels from a small number of viewing directions(typically3to6CCD cameras)and secondly it involves high spatial frequencies.In these conditions algebraic reconstruction methods are more appropriate with respect to analytical reconstruction methods,such as Fourier and back-projection methods (Verhoeven1993).For this reason,only the former class of methods is considered for further evaluation.4.1Algebraic methodsAlgebraic methods(Herman and Lent1976)iteratively solve a set of linear equations modeling the imaging system.In the present approach the measurement volume containing the particle distribution(the object) is discretized as a3D array of cubic voxel elements in (X,Y,Z)(in tomography referred to as the basis functions)with intensity E(X,Y,Z).A cubic voxel element has a uniform non-zero value inside and zero outside and its size is chosen comparable to that of a pixel,because particle images need to be properly discretized in the object as it is done in the images. Moreover,the interrogation by cross-correlation can be easily extended from a pixel to a voxel based object. Then the projection of the light intensity distribution E(X,Y,Z)onto an image pixel(x i,y i)returns the pixel intensity I(x i,y i)(known from the recorded images), which is written as a linear equation:Xj2N iw i;j EðX j;Y j;Z jÞ¼Iðx i;y iÞ;ð1Þwhere N i indicates the voxels intercepted or in the neighborhood of the line of sight corresponding to the i th pixel(x i,y i)(shaded voxels in Fig.2).The weighting coefficient w i,j describes the contribution of the j th voxel with intensity E(X j,Y j,Z j)to the pixel intensity I(x i,y i)and is calculated as the intersecting volume between the voxel and the line of sight(having the cross sectional area of the pixel)normalized with the voxel volume.The coefficients depend on the relative size of a voxel to a pixel and the distance between the voxel center and the line of sight(distance d in Fig.2). Note that0£w i,j£1for all entries w i,j in the2D array W and that W is very sparse,since a line of sight intersects with only a small part of the total volume. The weighting coefficients can also be used to account for different camera sensitivities,forward or backward scatter differences or other optical dissimilarities be-tween the cameras.Alternatively the recorded images can be pre-process in an appropriate way to compen-sate for these effects.In the above model,applying geometrical optics,the recorded pixel intensity is the object intensity E(X,Y, Z)integrated along the corresponding line of sight.In that case the reconstructed particle is represented by a 3D Gaussian-type blob,which projection in all direc-tions is the diffraction spot.A range of algebraic tomographic reconstruction algorithms is available to solve these equations.How-ever,due to the nature of the system,the problem is underdetermined(Sect.3)and the calculation may converge to different solutions,which implies that these algorithms solve the set of equations of Eq.1 under different optimization criteria.A detailed dis-cussion and analysis of these optimization criteria isbeyond the scope of the present study and can be found in Herman and Lent(1976).Instead the per-formance of two different tomographic algorithms is evaluated by numerical simulations of a tomographic-PIV experiment focusing the evaluation upon the reconstruction accuracy and convergence properties. The comparison is performed between the additive and multiplicative techniques referred to as ART(alge-braic reconstruction technique)and MART(multipli-cative algebraic reconstruction technique),respectively (Herman and Lent1976).Starting from a suitable ini-tial guess(E(X,Y,Z)0is uniform,see next section)the object E(X,Y,Z)is updated in each full iteration as:1.for each pixel in each camera i:2.for each voxel j:ART:EðX j;Y j;Z jÞkþ1¼EðX j;Y j;Z jÞkþl Iðx i;y iÞÀPj2N iw i;j EðX j;Y j;Z jÞkPj2N i i;jw i;jð2ÞMART:EðX j;Y j;Z jÞkþ1¼EðX j;Y j;Z jÞk Iðx i;y iÞ,Xj2N iw i;j EðX j;Y j;Z jÞk!l w i;jð3Þend loop2end loop1where l is a scalar relaxation parameter,which forART is between0and2and for MART must be£1.In ART the magnitude of the correction depends on the residual Iðx i;y iÞÀPj2N iw i;j EðX j;Y j;Z jÞmultiplied by a scaling factor and the weighting coefficient,so that only the elements in E(X,Y,Z)affecting the i th pixel are updated.Alternatively in MART the magnitude of the update is determined by the ratio of the measuredpixel intensity I with the projection of the current object Pj2N iw i;j EðX j;Y j;Z jÞThe exponent again ensures that only the elements in E(X,Y,Z)affecting the i th pixel are updated.Furthermore,the multiplicative MART scheme requires that E and I are definite positive.4.22-D numerical assessmentA domain with reduced dimensionality was adopted to evaluate the performances of the two methods:a 50·10mm22D slice of the3D volume.The2D par-ticlefield contains50particles,which images are re-corded by three linear array cameras with1,000pixel. For the ART reconstruction the relaxation parameter is set at0.2and the initial condition is a uniform zero intensity distribution,while for MART relaxation and initial condition are both1.The reconstructed particlefields after20iterations are shown in Fig.3.The maximum intensity is75 counts and values below3counts are blanked for readability.The ART algorithm leaves traces of the particles along the lines of sight,while the MART reconstruction shows more distinct particles.Theadditive ART scheme works similar to an OR-opera-tor:in order to have a non-zero intensity in the object it is sufficient to have a particle at the corresponding location in one of the PIV recordings.Still the intensity is highest at the actual particle locations.The multi-plicative MART scheme behaves as AND-operator:non-zero intensity only at locations where a particle appears in all recordings.The suitability of MART to reconstruct objects with sharp gradients or spikes has been confirmed by Verhoeven (1993).In conclusion,the artifacts in the ART reconstruction are undesir-able,especially considering that the signal has to be cross-correlated for interrogation to provide the de-sired velocity (displacement)information.The result from the MART algorithm better suits tomographic-PIV.Moreover,Fig.4shows that the individual parti-cles reconstructed with MART are reconstructed at the correct position with an intensity distribution slightly elongated in depth (d z ~1.5d x for the present case).Besides the sequential update algorithms of Eqs.2and 3schemes that update the reconstructed object simultaneously at every pixel exist,using all equations in a single step,such as the conjugate gradient method (additive scheme)as well as several implementations of the simultaneous MART algorithm (Mishra et al.1999).Those methods return similar results and are potentially computationally more efficient.The investigation of those methods goes beyond the scope of the present study and may be the subject of further investigations.5Parametric studyThis section discusses the effect of experimental and algorithm parameters on the reconstruction based onnumerical simulations.The number of iterations,the number of cameras,the viewing directions,the particle image density,the calibration accuracy and the image noise are considered.In the simulation the order of the problem is re-duced from a 3D volume with 2D images to a 2D slice with 1D images,which simplifies the computation and the interpretation of the results without loosing gen-erality on the results.In fact the 2D volume can be seen as a single slice selected from a 3D volume and similarly the 1D image as a single row of pixels taken from a 2D image.Tracer particles are distributed in a 50·10mm 2slice,which is imaged along a line of 1,000pixels from different viewing directions h (Fig.2)by cameras placed at infinity,such that magnification and viewing direction are constant over the field of view and the magnification is identical for all views and close to 1.Furthermore,the entire volume is assumed to be within the focal depth.Given the optical arrangement the particle location in the images is calculated and the application of diffraction (particle diameter is 3pixels,which is justified by the large f /#required)results in the synthetic recordings.The particle image intensity is assumed uniform.The 2D particle field is recon-structed from these recordings at 1,000·200voxel resolution using the MART algorithm described in the previous section.Unless stated otherwise the following (baseline)experimental parameters are assumed:three cameras at h ={–20,0,20}degrees,50particles (0.05particles per pixel),5reconstruction iterations and no calibration errors or image noise.To quantify the accuracy of the reconstruction pro-cess,the reconstructed object E 1(X ,Z )is compared with the exact distribution of light intensity E 0(X ,Z )-25-20-15-10-505101520250510ARTx (mm)0204060-25-20-15-10-505101520250510MART204060where the particles are represented by a Gaussian intensity distribution of three voxel diameter.The reconstruction quality Q is defined as the normalized correlation coefficient of the exact and reconstructed intensity distribution,according to:Q ¼PX ;ZE 1ðX ;Z ÞÁE 0ðX ;Z ÞffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiP X ;ZE 21ðX ;Z ÞÁP X ;ZE 20ðX ;Z Þr :ð4ÞA direct estimate of the correlation coefficient to beexpected in the cross-correlation analysis of the reconstructed objects is therefore given by Q 2in the assumption of perfect cross-correlation of the corre-sponding Gaussian particle fields and uncorrelated reconstruction noise.First the convergence of the residual and the reconstruction is considered.The residual (Fig.5,left)decreases monotonically in the tomographic iterative process.Although this behavior does not guarantee that the reconstructed object converges to the exact solution (Watt and Conery 1993),for the present reconstruction algorithm and noise free recordings,the solution converges correctly as evident from Fig.5,right.A small tendency to divergence after four to five iterations was observed only for a noise level in the recordings in excess of 50%with respect to the particle peak intensity.Such divergence can be countered by applying a smaller relaxation (l =0.2)as suggested by Mishra et al.(1999),although that requires more iter-ations to reach a similar reconstruction quality Q .In the present parametric study and in the experiments presented in Sect.7the reconstruction process is stopped after five iterations since both the residual and reconstruction quality Q do not change significantly performing further iterations.Moreover,any diver-gence phenomenon is avoided.Furthermore,theexperimental results show that with additional itera-tions the returned vector field changes only within the noise level.Figure 6shows the dependence of the reconstruc-tion quality Q on the most relevant experimental parameters,namely the number of cameras,the view-ing angle,the particle image density (in particles per pixel),and the calibration error.The diagrams show clear trends,which provides the experimentalist with a first indication of the optimum experimental arrange-ment and the limitations of the system.A correlation coefficient of 75%with respect to the exact spatial distribution (Q =0.75corresponding to an expected cross-correlation coefficient of Q 2=0.56)is used as a cut-off value,above which the reconstruction should be considered sufficiently accurate.The effect of the number of cameras is clear:adding camera gives additional information on the object,which increases reconstruction accuracy.A 2-camera system (h =–20°and 20°)is largely insufficient,whereas Q rapidly increases going to three and four cameras (h =–20°,0°,20°and 40°)and approaches unity with five cameras (h =–40°,–20°,0°,20°and 40°).The viewing angles are changed maintaining the symmetric camera arrangement.The angle indicated in Fig.6is the angle between the outer cameras and the z -axis.The graph shows an optimum near 30°.For smaller angles the depth resolution decreases resulting in elongation of the reconstructed-particle in depth.For larger angles the intercepted length of the line-of-sight increases,which causes a larger number of par-ticle to be formed with respect to those actually present in the illuminated volume.Such extra particles will be referred to as ‘‘ghost particles’’(Maas et al.1993).This is a problem of ambiguity,which increases with the number of particles,the particle diameter and the length of the line of sight in the volume.The latter increases with the viewing angle in the present con-figuration,hence the increase in ghost particles.The configurations returning an optimum have a viewing angle in the range of 15–45°.An increased particle density produces a larger percentage of ghost particles consequently decreasing the reconstruction quality.On the other hand,the larger number of particles allows a higher spatial sampling rate of the flow,returning a potentially higher spatial resolution.Therefore,a high particle density is desirable.Based on the simulation results (Fig.6)the maximum imaged particle density yielding an accept-able reconstruction quality is 0.075and 0.15particles per pixel for the three and four camera system,respectively.In a 3D system this would approximately102030405060。

Unit 34 Oil Pollution Responding and Damage Control油污染应急和损害控制1. Fate and behavior of oil in the marine environment石油在海洋环境中的演变和特性Complex processes of oil transformation in the marine environment start developing from the first seconds of oil’s contact with seawater. The progression, duration, and result of these transformations depend on the properties and composition of the oil itself, parameters of the actual oil spill, and environmental conditions. The main characteristics of oil transformations are their dynamism, especially at the first stages, and the close interaction of physical, chemical, and biological mechanisms of dispersion and degradation intoxicated living organism, a marine ecosystem destroys, metabolizes, and deposits the excessive amounts of hydrocarbons, transforming them into more common and safer substances.自从石油及其制品与海洋接触的那一该开始，在海洋环境中转化石油的复杂性就在不断增长。

第３８卷第１８期２０１８年９月生态学报ＡＣＴＡＥＣＯＬＯＧＩＣＡＳＩＮＩＣＡＶｏｌ．３８，Ｎｏ．１８Ｓｅｐ．，２０１８基金项目：国家自然科学基金项目（４１６７１０２７，４１７３０７４９）收稿日期：２０１８⁃０２⁃２２；㊀㊀修订日期：２０１８⁃０７⁃０２∗通讯作者Ｃｏｒｒｅｓｐｏｎｄｉｎｇａｕｔｈｏｒ．Ｅ⁃ｍａｉｌ：ｍａｙｉｎｇ＠ｉｇｓｎｒｒ．ａｃ．ｃｎＤＯＩ：１０．５８４６／ｓｔｘｂ２０１８０２２５０３８４杜俊杉，马英，胡晓农，童菊秀，张宝忠，孙宁霞，高光耀．基于双稳定同位素和ＭｉｘＳＩＡＲ模型的冬小麦根系吸水来源研究．生态学报，２０１８，３８（１８）：６６１１⁃６６２２．ＤｕＪＳ，ＭａＹ，ＨｕＸＮ，ＴｏｎｇＪＸ，ＺｈａｎｇＢＺ，ＳｕｎＮＸ，ＧａｏＧＹ．ＡｐｐｌｙｉｎｇｄｕａｌｓｔａｂｌｅｉｓｏｔｏｐｅｓａｎｄａＭｉｘＳＩＡＲｍｏｄｅｌｔｏｄｅｔｅｒｍｉｎｅｒｏｏｔｗａｔｅｒｕｐｔａｋｅｏｆｗｉｎｔｅｒｗｈｅａｔ．ＡｃｔａＥｃｏｌｏｇｉｃａＳｉｎｉｃａ，２０１８，３８（１８）：６６１１⁃６６２２．基于双稳定同位素和ＭｉｘＳＩＡＲ模型的冬小麦根系吸水来源研究杜俊杉１，马㊀英２，∗，胡晓农１，童菊秀１，张宝忠３，孙宁霞１，高光耀４１中国地质大学（北京）水资源与环境学院，北京㊀１０００８３２中国科学院地理科学与资源研究所，陆地水循环及地表过程院重点实验室，北京㊀１００１０１３中国水利水电科学研究院，流域水循环模拟与调控国家重点实验室，北京㊀１０００３８４中国科学院生态环境研究中心，城市与区域生态国家重点实验室，北京㊀１０００８５摘要：灌溉和施肥措施对农田水文循环具有重要影响，根系吸水是联系植物蒸腾和土壤水分运动的关键水文过程，定量识别灌溉施肥影响下作物根系吸水来源对农业用水优化管理具有重要意义㊂氘氧稳定同位素（Ｄ和１８Ｏ）是追溯农田水分运移过程的理想天然示踪剂㊂基于２０１３ ２０１５年北京市典型农田不同灌溉施肥处理冬小麦水分运移试验，利用Ｄ和１８Ｏ双稳定同位素和ＭｉｘＳＩＡＲ贝叶斯混合模型，量化冬小麦主要根系吸水深度及其贡献比例，阐明作物水分来源的季节变化及不同处理间的差异，分析根系吸水与土壤水分分布变化的相互关系㊂研究结果表明：两季冬小麦返青⁃拔节㊁拔节⁃抽穗㊁抽穗⁃灌浆和灌浆⁃收获期主要根系吸水深度均值分别为０ ２０ｃｍ（６７．０％）㊁２０ ７０ｃｍ（４２．０％）㊁０ ２０ｃｍ（３８．７％）和２０ ７０ｃｍ（３４．９％），但季节变化差异显著，２０１４季主要吸水深度随作物的生长发育而逐渐增加，２０１５季则主要集中于浅层土壤（０ ７０ｃｍ）㊂返青⁃抽穗期仅灌水２０ｍｍ或施肥１０５ｋｇ／ｈｍ２Ｎ促使拔节⁃抽穗期深层（７０ ２００ｃｍ）土壤水分利用率平均增加２９％，而前期充分灌水且大量施肥（ȡ当地施肥量２１０ｋｇｈｍ－２Ｎ）时拔节⁃抽穗期根系吸水深度为土壤表层０ ２０ｃｍ㊂在干旱少雨的冬小麦生长季内作物吸水来源与土壤水分消耗变化基本一致㊂关键词：双稳定同位素；根系吸水；ＭｉｘＳＩＡＲ模型；灌溉施肥处理；冬小麦ＡｐｐｌｙｉｎｇｄｕａｌｓｔａｂｌｅｉｓｏｔｏｐｅｓａｎｄａＭｉｘＳＩＡＲｍｏｄｅｌｔｏｄｅｔｅｒｍｉｎｅｒｏｏｔｗａｔｅｒｕｐｔａｋｅｏｆｗｉｎｔｅｒｗｈｅａｔＤＵＪｕｎｓｈａｎ１，ＭＡＹｉｎｇ２，∗，ＨＵＸｉａｏｎｏｎｇ１，ＴＯＮＧＪｕｘｉｕ１，ＺＨＡＮＧＢａｏｚｈｏｎｇ３，ＳＵＮＮｉｎｇｘｉａ１，ＧＡＯＧｕａｎｇｙａｏ４１ＳｃｈｏｏｌｏｆＷａｔｅｒＲｅｓｏｕｒｃｅｓａｎｄＥｎｖｉｒｏｎｍｅｎｔ，ＣｈｉｎａＵｎｉｖｅｒｓｉｔｙｏｆＧｅｏｓｃｉｅｎｃｅｓ（Ｂｅｉｊｉｎｇ），Ｂｅｉｊｉｎｇ１０００８３，Ｃｈｉｎａ２ＫｅｙＬａｂｏｒａｔｏｒｙｏｆＷａｔｅｒＣｙｃｌｅａｎｄＲｅｌａｔｅｄＬａｎｄＳｕｒｆａｃｅＰｒｏｃｅｓｓｅｓ，ＩｎｓｔｉｔｕｔｅｏｆＧｅｏｇｒａｐｈｉｃＳｃｉｅｎｃｅｓａｎｄＮａｔｕｒａｌＲｅｓｏｕｒｃｅｓＲｅｓｅａｒｃｈ，ＣｈｉｎｅｓｅＡｃａｄｅｍｙｏｆＳｃｉｅｎｃｅｓ，Ｂｅｉｊｉｎｇ１００１０１，Ｃｈｉｎａ３ＳｔａｔｅＫｅｙＬａｂｏｒａｔｏｒｙｏｆＳｉｍｕｌａｔｉｏｎａｎｄＲｅｇｕｌａｔｉｏｎｏｆＷａｔｅｒＣｙｃｌｅｉｎＲｉｖｅｒＢａｓｉｎ，ＣｈｉｎａＩｎｓｔｉｔｕｔｅｏｆＷａｔｅｒＲｅｓｏｕｒｃｅｓａｎｄＨｙｄｒｏｐｏｗｅｒＲｅｓｅａｒｃｈ，Ｂｅｉｊｉｎｇ１０００３８，Ｃｈｉｎａ４ＳｔａｔｅＫｅｙＬａｂｏｒａｔｏｒｙｏｆＵｒｂａｎａｎｄＲｅｇｉｏｎａｌＥｃｏｌｏｇｙ，ＲｅｓｅａｒｃｈＣｅｎｔｅｒｆｏｒＥｃｏ⁃ＥｎｖｉｒｏｎｍｅｎｔａｌＳｃｉｅｎｃｅｓ，ＣｈｉｎｅｓｅＡｃａｄｅｍｙｏｆＳｃｉｅｎｃｅｓ，Ｂｅｉｊｉｎｇ１０００８５，ＣｈｉｎａＡｂｓｔｒａｃｔ：Ｉｒｒｉｇａｔｉｏｎａｎｄｆｅｒｔｉｌｉｚａｔｉｏｎｓｃｈｅｄｕｌｉｎｇｓｉｇｎｉｆｉｃａｎｔｌｙａｆｆｅｃｔｆｉｅｌｄｗａｔｅｒｃｙｃｌｅｓ．Ｒｏｏｔｗａｔｅｒｕｐｔａｋｅｉｓｔｈｅｃｒｉｔｉｃａｌｐｒｏｃｅｓｓｔｈａｔｃｏｎｎｅｃｔｓｐｌａｎｔｔｒａｎｓｐｉｒａｔｉｏｎａｎｄｓｏｉｌｗａｔｅｒｍｏｖｅｍｅｎｔｓ．Ｉｔｉｓｎｅｃｅｓｓａｒｙｔｏｑｕａｎｔｉｆｙｔｈｅｃｒｏｐｗａｔｅｒｕｐｔａｋｅｓｏｕｒｃｅｓ２１６６㊀生㊀态㊀学㊀报㊀㊀㊀３８卷㊀ｕｎｄｅｒｄｉｆｆｅｒｅｎｔｉｒｒｉｇａｔｉｏｎａｎｄｆｅｒｔｉｌｉｚａｔｉｏｎｔｒｅａｔｍｅｎｔｓｆｏｒｏｐｔｉｍｉｚｉｎｇａｇｒｉｃｕｌｔｕｒａｌｗａｔｅｒｍａｎａｇｅｍｅｎｔ．ＴｈｅｓｔａｂｌｅｗａｔｅｒｉｓｏｔｏｐｅｓｏｆＤａｎｄ１８Ｏａｒｅｃｏｎｓｉｄｅｒｅｄａｓｉｄｅａｌ（ｎａｔｕｒａｌ）ｔｒａｃｅｒｓｆｏｒｔｒａｃｋｉｎｇｗａｔｅｒｔｈｒｏｕｇｈｔｈｅｓｏｉｌｂａｓｅｄｏｎｄｉｓｔｉｎｃｔｉｓｏｔｏｐｉｃｓｉｇｎａｔｕｒｅｓｏｆｗａｔｅｒｆｌｕｘｅｓ．ＴｈｅＭｉｘＳＩＡＲｆｒａｍｅｗｏｒｋｉｓｔｈｅｌａｔｅｓｔＢａｙｅｓｉａｎｓｔａｂｌｅｉｓｏｔｏｐｅａｎａｌｙｓｉｓｍｉｘｉｎｇｍｏｄｅｌｉｎＲｔｈａｔｃｏｎｓｉｄｅｒｓｍｕｌｔｉｐｌｅｓｏｕｒｃｅｓｏｆｕｎｃｅｒｔａｉｎｔｙａｎｄｐｒｏｖｉｄｅｓｄｅｆｉｎｉｔｅｐｒｏｐｏｒｔｉｏｎｓｏｆｓｏｕｒｃｅｃｏｎｔｒｉｂｕｔｉｏｎｓ．Ｉｎｔｈｉｓｓｔｕｄｙ，ｄｕａｌｓｔａｂｌｅｉｓｏｔｏｐｅｓｏｆｓｏｉｌｗａｔｅｒａｎｄｓｔｅｍｗａｔｅｒｗｅｒｅａｐｐｌｉｅｄｔｏｄｅｔｅｒｍｉｎｅｓｅａｓｏｎａｌｗａｔｅｒｕｐｔａｋｅｐａｔｔｅｒｎｓｏｆｗｉｎｔｅｒｗｈｅａｔｕｎｄｅｒｄｉｆｆｅｒｅｎｔｉｒｒｉｇａｔｉｏｎａｎｄｆｅｒｔｉｌｉｚａｔｉｏｎｔｒｅａｔｍｅｎｔｓｄｕｒｉｎｇ２０１３ ２０１５ｉｎＢｅｉｊｉｎｇ，Ｃｈｉｎａ．ＴｈｅｃｏｎｔｒｉｂｕｔｉｏｎｓｏｆｔｈｅｓｏｉｌｗａｔｅｒａｔｅａｃｈｄｅｐｔｈｔｏｗａｔｅｒｕｐｔａｋｅｗｅｒｅｑｕａｎｔｉｆｉｅｄｕｓｉｎｇｔｈｅＭｉｘＳＩＡＲＢａｙｅｓｉａｎｍｉｘｉｎｇｍｏｄｅｌ．Ｔｈｅｄｉｒｅｃｔｉｎｆｅｒｅｎｃｅｍｅｔｈｏｄｗａｓａｌｓｏｕｓｅｄｔｏｄｅｔｅｃｔｔｈｅｐｏｔｅｎｔｉａｌｒｏｏｔｗａｔｅｒｕｐｔａｋｅｄｅｐｔｈ．Ｃｏｒｒｅｌａｔｉｏｎｓｂｅｔｗｅｅｎｔｈｅｗａｔｅｒｕｐｔａｋｅｐａｔｔｅｒｎｓａｎｄｓｏｉｌｍｏｉｓｔｕｒｅｃｈａｎｇｅｓｗｅｒｅｆｕｒｔｈｅｒｅｖａｌｕａｔｅｄ．Ｔｈｅａｖｅｒａｇｅｃｏｎｔｒｉｂｕｔｉｏｎｏｆｓｏｉｌｗａｔｅｒｉｎｔｈｅ０ ２０，２０ ７０，７０ １５０，ａｎｄ１５０ ２００ｃｍｌａｙｅｒｓｗａｓ３５．６％，２７．６％，２３．１％，ａｎｄ１３．７％，ｒｅｓｐｅｃｔｉｖｅｌｙ．Ｔｈｅｐｒｉｍａｒｙｒｏｏｔｗａｔｅｒｕｐｔａｋｅｄｅｐｔｈｗａｓ０ ２０ｃｍ（６７．０％），２０ ７０ｃｍ（４２．０％），０ ２０ｃｍ（３８．７％），ａｎｄ２０ ７０ｃｍ（３４．９％）ｄｕｒｉｎｇｔｈｅｇｒｅｅｎｉｎｇ⁃ｊｏｉｎｔｉｎｇ，ｊｏｉｎｔｉｎｇ⁃ｈｅａｄｉｎｇ，ｈｅａｄｉｎｇ⁃ｆｉｌｌｉｎｇ，ａｎｄｆｉｌｌｉｎｇ⁃ｈａｒｖｅｓｔｐｅｒｉｏｄｓ，ｒｅｓｐｅｃｔｉｖｅｌｙ．Ｓｉｇｎｉｆｉｃａｎｔｄｉｆｆｅｒｅｎｃｅｓｉｎｃｒｏｐｗａｔｅｒｕｓｅａｐｐｅａｒｅｄｂｅｔｗｅｅｎｔｈｅ２０１４ａｎｄ２０１５ｇｒｏｗｉｎｇｓｅａｓｏｎｓ．Ｔｈｅｍａｉｎｒｏｏｔｗａｔｅｒｕｐｔａｋｅｄｅｐｔｈｇｒａｄｕａｌｌｙｉｎｃｒｅａｓｅｄｆｒｏｍ０ ２０ｃｍ（ｇｒｅｅｎｉｎｇ⁃ｊｏｉｎｔｉｎｇｐｅｒｉｏｄ）ｔｏ７０ １５０ｃｍ（ｈｅａｄｉｎｇ⁃ｆｉｌｌｉｎｇｐｅｒｉｏｄ）ａｎｄｗａｓｍａｉｎｔａｉｎｅｄａｔｔｈｅ７０ １５０ｃｍｄｅｐｔｈｕｎｔｉｌｔｈｅｆｉｌｌｉｎｇ⁃ｈａｒｖｅｓｔｐｅｒｉｏｄｉｎｔｈｅ２０１４ｓｅａｓｏｎ．Ｈｏｗｅｖｅｒ，ｗｉｎｔｅｒｗｈｅａｔｍａｉｎｌｙｔｏｏｋｕｐｓｏｉｌｗａｔｅｒｆｒｏｍｔｈｅｓｈａｌｌｏｗｌａｙｅｒｓ（０ ７０ｃｍ）ｏｖｅｒｔｈｅ２０１５ｓｅａｓｏｎ．Ｉｎｐａｒｔｉｃｕｌａｒ，ｔｈｅｐｒｏｐｏｒｔｉｏｎａｌｓｏｉｌｗａｔｅｒｃｏｎｔｒｉｂｕｔｉｏｎｉｎｔｈｅ０ ２０ｃｍｌａｙｅｒｗａｓｒｅｍａｒｋａｂｌｙｈｉｇｈｅｒ（１３．９％）ｔｈａｎｔｈａｔｉｎｔｈｅ２０１４ｓｅａｓｏｎ．Ｒｏｏｔｗａｔｅｒｕｐｔａｋｅｐａｔｔｅｒｎｓｗｉｔｈｓｏｉｌｍｏｉｓｔｕｒｅｄｉｓｔｒｉｂｕｔｉｏｎｓｗｅｒｅｓｉｇｎｉｆｉｃａｎｔｌｙｉｎｆｌｕｅｎｃｅｄｂｙｄｉｆｆｅｒｅｎｔｉｒｒｉｇａｔｉｏｎａｎｄｆｅｒｔｉｌｉｚａｔｉｏｎｔｒｅａｔｍｅｎｔｓ，ｅｓｐｅｃｉａｌｌｙｉｎｄｒｙｓｅａｓｏｎｓ．Ｔｈｅｍａｉｎｗａｔｅｒｕｐｔａｋｅｓｏｕｒｃｅｗａｓｔｈｅｓｏｉｌｗａｔｅｒｉｎｔｈｅｔｏｐｌａｙｅｒ（０ ２０ｃｍ）ｕｎｄｅｒｔｈｅＴ４ａｎｄＴ５ｔｒｅａｔｍｅｎｔｓｄｕｒｉｎｇｔｈｅｊｏｉｎｔｉｎｇ⁃ｈｅａｄｉｎｇｐｅｒｉｏｄｉｎ２０１５，ｂｅｃａｕｓｅｔｈｅｒｏｏｔｇｒｏｗｔｈｉｎｔｈｅｓｕｒｆａｃｅｌａｙｅｒｗａｓｓｔｉｍｕｌａｔｅｄｂｙｓｕｆｆｉｃｉｅｎｔｉｒｒｉｇａｔｉｏｎ（８０ｍｍ）ａｎｄａｂｕｎｄａｎｔｆｅｒｔｉｌｉｚａｔｉｏｎ（ȡ２１０ｋｇ／ｈｍ２Ｎ）ａｔｅａｒｌｙｇｒｏｗｔｈｓｔａｇｅｓ．Ｎｉｔｒｏｇｅｎｄｅｆｉｃｉｅｎｃｙｗｉｔｈ＜１０５ｋｇ／ｈｍ２Ｎ（Ｔ３）ｏｒｌｅｓｓｉｒｒｉｇａｔｉｏｎｗｉｔｈ２０ｍｍ（Ｔ１ａｎｄＴ２）ｄｕｒｉｎｇｔｈｅｇｒｅｅｎｉｎｇ⁃ｊｏｉｎｔｉｎｇｐｅｒｉｏｄｐｒｏｍｏｔｅｄｒｏｏｔｇｒｏｗｔｈｉｎｔｈｅｄｅｅｐｓｏｉｌｌａｙｅｒ（７０ ２００ｃｍ）ａｎｄｉｎｃｒｅａｓｅｄｗａｔｅｒａｄｓｏｒｐｔｉｏｎｂｙａｍｅａｎｏｆ２９％ｄｕｒｉｎｇｔｈｅｊｏｉｎｔｉｎｇ⁃ｆｉｌｌｉｎｇｐｅｒｉｏｄ．Ｓｅａｓｏｎａｌｖａｒｉａｔｉｏｎｓｉｎｔｈｅｑｕａｎｔｉｔａｔｉｖｅｃｏｎｔｒｉｂｕｔｉｏｎｏｆｓｏｉｌｗａｔｅｒａｔｄｉｆｆｅｒｅｎｔｄｅｐｔｈｓｗｅｒｅｃｌｏｓｅｌｙｒｅｌａｔｅｄｔｏｔｈｅｓｏｉｌｍｏｉｓｔｕｒｅｄｉｓｔｒｉｂｕｔｉｏｎｓ．Ｔｈｅｌａｒｇｅｃｏｎｔｒｉｂｕｔｉｏｎｏｆｓｏｉｌｗａｔｅｒｉｎｔｈｅ０ １５０ｃｍｌａｙｅｒ（８６．３％）ｗａｓｃｏｎｓｉｓｔｅｎｔｗｉｔｈｉｔｓｐｒｏｐｏｒｔｉｏｎａｌｃｏｎｓｕｍｐｔｉｏｎｉｎｓｏｉｌｗａｔｅｒｓｔｏｒａｇｅ（９２％）ｔｈｒｏｕｇｈｏｕｔｔｈｅｇｒｅｅｎｉｎｇｔｏｈａｒｖｅｓｔｓｅａｓｏｎｏｆｗｉｎｔｅｒｗｈｅａｔ．Ｔｈｉｓｓｔｕｄｙｐｒｏｖｉｄｅｓａｓｉｍｐｌｅａｎｄｅｆｆｅｃｔｉｖｅｍｅｔｈｏｄｆｏｒｉｄｅｎｔｉｆｙｉｎｇｃｒｏｐｗａｔｅｒｓｏｕｒｃｅｓ．Ｔｈｅｆｉｎｄｉｎｇｓａｒｅｏｆｇｒｅａｔｓｉｇｎｉｆｉｃａｎｃｅｆｏｒｆｕｔｕｒｅｆｅｒｔｉｌｉｚａｔｉｏｎａｎｄｉｒｒｉｇａｔｉｏｎｍａｎａｇｅｍｅｎｔ．ＫｅｙＷｏｒｄｓ：ｄｕａｌｓｔａｂｌｅｉｓｏｔｏｐｅｓ；ｒｏｏｔｗａｔｅｒｕｐｔａｋｅ；ＭｉｘＳＩＡＲｍｏｄｅｌ；ｉｒｒｉｇａｔｉｏｎａｎｄｆｅｒｔｉｌｉｚａｔｉｏｎｔｒｅａｔｍｅｎｔ；ｗｉｎｔｅｒｗｈｅａｔ灌溉和施肥措施对农田水文循环具有重要影响㊂根系吸水是联系植物蒸腾和土壤水分运动的关键过程［１］㊂定量识别不同灌溉施肥条件下植物根系吸水来源对农田水分高效利用具有重要意义㊂以往研究根系吸水多采用根系挖掘法，但该方法容易破坏植物生境［２⁃３］㊂稳定同位素（Ｄ和１８Ｏ）因其本身即为水分子组成部分，且在根系吸水过程中不发生分馏，是识别植物根系吸水来源的理想天然示踪剂［４⁃５］㊂利用Ｄ和１８Ｏ同位素进行植物水分溯源的方法主要包括：直接对比法［６］，两端元和三端元线性混合模型［７］，ＩｓｏＳｏｕｒｃｅ混合模型［８］，基于贝叶斯理论的ＭｉｘＳＩＲ㊁ＳＩＡＲ和ＭｉｘＳＩＡＲ混合模型［９⁃１１］等㊂直接对比法假设植物仅利用某一深度的土壤水，而实际上植物水分通常是多个深度土壤水的混合［２，１２⁃１３］㊂端元混合模型则局限于计算三种及以下水分来源对植物水的贡献比例［１４⁃１５］㊂ＩｓｏＳｏｕｒｃｅ模型可估算多种水分来源对植物水的贡献比例范围［１，２，１６］，但忽略了水源的不确定性且其贡献比例范围不能以单一概率表示㊂基于贝叶斯统计原理的混合模型则充分考虑了混合物（植物茎水）和贡献源（土壤水）同位素值的潜在不确定性以及贡献源过参数化导致的不确定性，其中ＭｉｘＳＩＡＲ模型最新融合了ＭｉｘＳＩＲ和ＳＩＡＲ模型的优势，增加了贡献源的多元同位素原始数据源输入形式㊁随机效应分类变量和残差＋过程误差等模块，显著提升了植物水分来源及其贡献比例定量计算的准确性［１１］㊂伴随同位素测试技术的迅速发展，基于Ｄ和１８Ｏ同位素的植物水分溯源方法被广泛用于农业系统中［１⁃３，１７⁃２３］㊂Ｚｈａｎｇ等［１７］利用两端元线性混合模型和δＤ δ１８Ｏ关系图计算了不同生育阶段夏玉米的水分来源，但该模型需要所有水分来源的δＤ和δ１８Ｏ之间具有强线性相关关系㊂其中，应用较为普遍的ＩｓｏＳｏｕｒｃｅ混合模型被用于估算美国Ｉｏｗａ中部地区玉米㊁华北地下水浅埋区夏玉米和棉花和地下水深埋区充分灌溉条件下冬小麦㊁西北地区石羊河流域交替隔沟灌溉条件下玉米以及西北内陆干旱区（新疆）棉花的主要根系吸水深度［１⁃３，１８⁃２０］㊂安江龙等［２１］采用底部隔水的塑料管土柱在田间进行了冬小麦种植试验，并应用耦合模型和ＩｓｏＳｏｕｒｃｅ模型对比分析了不同生育阶段水源贡献率㊂相比之下，贝叶斯混合模型利用统计分布方法刻画了作物水分溯源过程中的不确定性，其计算结果更加准确可靠㊂Ｙａｎｇ等［２２］采用ＭｉｘＳＩＲ模型确定了黑河流域中游地区玉米的吸水深度，结果表明当地农业灌溉用水效率较低，灌水深度远大于作物根系吸水深度；Ｍａ和Ｓｏｎｇ［２３］基于ＭｉｘＳＩＡＲ模型对比分析了不同施肥条件下夏玉米根系吸水深度的变化规律㊂但利用Ｄ和１８Ｏ同位素和ＭｉｘＳＩＡＲ模型定量识别灌溉和施肥耦合作用下作物吸水来源季节变化的研究尚未见报道㊂为此，本文基于北京市典型农田不同灌溉施肥条件下冬小麦水分运移试验，采用Ｄ和１８Ｏ双稳定同位素和ＭｉｘＳＩＡＲ模型量化不同深度土壤水分对作物水的贡献比例，对比不同灌溉施肥条件下冬小麦根系吸水来源的季节分布差异，探讨根系吸水深度与土壤水分变化的关系，该研究可为农田水分高效利用提供科学指导㊂１㊀材料与方法１．１㊀田间试验试验于２０１３ ２０１５年在国家节水灌溉工程技术研究中心（北京）大兴试验研究基地（３９ʎ３７ᶄＮ，１１６ʎ２６ᶄＥ）进行㊂研究区属半湿润大陆性季风气候，多年平均降水量５４０ｍｍ，但７０％ ８０％的降雨集中在夏季，年平均气温１２．１ħ［２４］㊂试验基地０ ２００ｃｍ土壤剖面的物理和化学特性参数见表１㊂地下水位埋深约１６ｍ㊂冬小麦供试品种为中麦１７５，分别于２０１３年和２０１４年１０月初播种，次年６月初收获㊂试验共设计５个水肥耦合处理，灌溉施肥方案见表２，其中Ｔ５处理参考当地常规灌溉施肥水平，总灌溉量和施肥量分别为３００ｍｍ和３１５ｋｇ／ｈｍ２Ｎ㊂试验在小区内进行，每个小区面积约６ｍˑ５ｍ，每个处理３组重复㊂两季小麦返青后的第二次灌水时间不同，２０１４季在拔节⁃抽穗期，２０１５季在抽穗⁃灌浆期㊂灌溉水源为当地地下水，灌溉方式为地面灌溉㊂小麦返青后追施尿素，施肥日期分别为２０１４年３月２７日和２０１５年３月２９日㊂土壤剖面含水量通过ＴＲＩＭＥ⁃ＩＰＨ土壤水分仪（ＩＭＫＯ，德国）测定，测量间距２０ｃｍ，每５ ７天测量一次，灌溉和降雨后加测㊂气象数据由站内的自动气象站（ＭｏｎｉｔｏｒＳｅｎｓｏｒｓ，Ａｕｓｔｒａｌｉａ）监测，主要监测指标包括日降水量㊁日最高和最低温度㊁风速㊁平均温度㊁平均相对湿度等㊂２０１４季冬小麦生长季内降水量㊁平均温度和平均相对湿度分别为８５．９ｍｍ，７．３ħ和５３．１％；２０１５季则分别为８７．９ｍｍ，７．５ħ和４８．７％㊂２０１４和２０１５季参考作物蒸散发量分别为４３５ｍｍ和４５８ｍｍ㊂表１㊀土壤剖面（０ ２００ｃｍ）物理和化学特性参数Ｔａｂｌｅ１㊀Ｐｈｙｓｉｃａｌａｎｄｃｈｅｍｉｃａｌｐｒｏｐｅｒｔｉｅｓｏｆｔｈｅｓｏｉｌｐｒｏｆｉｌｅ深度Ｄｅｐｔｈ／ｃｍ土壤颗粒组成Ｐａｒｔｉｃｌｅｓｉｚｅ／％砂粒Ｓａｎｄ粉粒Ｓｉｌｔ粘粒Ｃｌａｙ土壤质地Ｓｏｉｌｔｅｘｔｕｒｅ干容重Ｂｕｌｋｄｅｎｓｉｔｙ／（ｇ／ｃｍ３）饱和含水量θｓＳａｔｕｒａｔｅｄｗａｔｅｒｃｏｎｔｅｎｔ／（ｃｍ３／ｃｍ３）饱和导水率ＫｓＳａｔｕｒａｔｅｄｈｙｄｒａｕｌｉｃｃｏｎｄｕｃｔｉｖｉｔｙ／（ｃｍ／ｄ）电导率ＥＣＥｌｅｃｔｒｉｃａｌｃｏｎｄｕｃｔｉｖｉｔｙ／（μＳ／ｃｍ）ｐＨ铵态氮ＮＨ＋４⁃Ｎ／（ｍｇ／ｋｇ）硝态氮ＮＯ－３⁃Ｎ／（ｍｇ／ｋｇ）㊀０ ２０５８．８３３．２８．０砂壤土１．５６０．４１８．４１１１１．７０８．１５７．０９８．９㊀２０ １２０６５．３２６．７８．０砂壤土１．４８０．４２１０．０４１０９．１２８．６１５．９１７．９１２０ １８０６８．２２９．２２．７砂壤土１．４５０．４５７．４５８７．６０８．６６６．３２０．２１８０ ２００３２．０５１．０１７．０粉壤土１．２５０．５１０．６６１６１．８０８．２７４．４１９．０３１６６㊀１８期㊀㊀㊀杜俊杉㊀等：基于双稳定同位素和ＭｉｘＳＩＡＲ模型的冬小麦根系吸水来源研究㊀表２㊀冬小麦灌溉施肥方案Ｔａｂｌｅ２Ｉｒｒｉｇａｔｉｏｎａｎｄｆｅｒｔｉｌｉｚａｔｉｏｎｓｃｈｅｄｕｌｅｏｆｅａｃｈｔｒｅａｔｍｅｎｔｆｏｒｗｉｎｔｅｒｗｈｅａｔ年份Ｙｅａｒ处理Ｔｒｅａｔｍｅｎｔ越冬⁃返青期Ｗｉｎｔｅｒｄｏｒｍａｎｃｙ⁃ｇｒｅｅｎｉｎｇ返青⁃拔节期Ｇｒｅｅｎｉｎｇ⁃ｊｏｉｎｔｉｎｇ拔节⁃抽穗期Ｊｏｉｎｔｉｎｇ⁃ｈｅａｄｉｎｇ抽穗⁃灌浆期Ｈｅａｄｉｎｇ⁃ｆｉｌｌｉｎｇ灌浆⁃收获期Ｆｉｌｌｉｎｇ⁃ｈａｒｖｅｓｔ总量Ｔｏｔａｌ灌溉Ｉｒｒｉｇａｔｉｏｎ／ｍｍ施肥Ｆｅｒｔｉｌｉｚａｔｉｏｎ／（ｋｇ／ｈｍ２Ｎ）灌溉Ｉｒｒｉｇａｔｉｏｎ／ｍｍ施肥Ｆｅｒｔｉｌｉｚａｔｉｏｎ／（ｋｇ／ｈｍ２Ｎ）灌溉Ｉｒｒｉｇａｔｉｏｎ／ｍｍ施肥Ｆｅｒｔｉｌｉｚａｔｉｏｎ／（ｋｇ／ｈｍ２Ｎ）灌溉Ｉｒｒｉｇａｔｉｏｎ／ｍｍ施肥Ｆｅｒｔｉｌｉｚａｔｉｏｎ／（ｋｇ／ｈｍ２Ｎ）灌溉Ｉｒｒｉｇａｔｉｏｎ／ｍｍ施肥Ｆｅｒｔｉｌｉｚａｔｉｏｎ／（ｋｇ／ｈｍ２Ｎ）灌溉Ｉｒｒｉｇａｔｉｏｎ／ｍｍ施肥Ｆｅｒｔｉｌｉｚａｔｉｏｎ／（ｋｇ／ｈｍ２Ｎ）２０１４Ｔ１６０－２０１０５８０－－－－－１６０１０５Ｔ２６０－２０３１５８０－－－－－１６０３１５Ｔ３６０－８０１０５－－－－８０－２２０１０５Ｔ４６０－８０３１５－－－－８０－２２０３１５Ｔ５６０－８０２１０８０－－－８０－３００２１０２０１５Ｔ１６０－２０１０５－－８０－－－１６０１０５Ｔ２６０－２０３１５－－８０－－－１６０３１５Ｔ３６０－８０１０５－－－－８０－２２０１０５Ｔ４６０－８０３１５－－－－８０－２２０３１５Ｔ５６０－８０２１０－－８０－８０－３００２１０㊀㊀Ｔ１：处理１，Ｔ２：处理２；Ｔ３：处理３；Ｔ４：处理４；Ｔ５：处理５；Ｔ５代表当地常规施肥水平３１５ｋｇ／ｈｍ２Ｎ； － 表示没有灌溉或施肥１．２㊀样品采集与同位素分析本研究于冬小麦返青至收获期间分别采集降水㊁灌溉水㊁土壤水和茎秆水并测定不同水体Ｄ和１８Ｏ同位素组成㊂次降水结束后通过由聚乙烯瓶和漏斗组成的收集器采集降水样［２］，并在漏斗口放置乒乓球以防止水面蒸发，收集的水样迅速转移到５０ｍＬ聚乙烯塑料瓶中并用封口膜密封㊂每次灌水时用聚乙烯塑料瓶采集灌溉水５０ｍＬ并密封保存㊂土壤水样通过土壤溶液提取器采集，采集深度１０，２０，３０，５０，７０，９０，１１０，１５０ｃｍ和２００ｃｍ，平均每周采集一次，灌溉和降雨之后加采㊂当土壤含水量较低无法获取土壤水时，采用土钻钻取土壤样品并迅速装入样品瓶中密封冷冻保存㊂采集土壤水和土壤样的同时，在其采样点周边选取代表性植株３个，每个处理３组重复，每个植株剪取土壤表面以上和第一个茎节以下的茎秆，迅速去除表皮后放入棕色玻璃样品瓶中，并用封口膜密封㊂所有的茎秆样品和土壤样品均在采集后立即置于－１５ħ到－２０ħ下冷冻保存㊂土壤和茎秆中的水分采用低温真空蒸馏系统（ＬＩ⁃２０００，ＬＩＣＡ，中国）提取㊂在中国科学院地理科学与资源研究所陆地水循环及地表过程院重点实验室采用ＬＧＲ液态水同位素分析仪（ＭｏｄｅｌＤＬＴ⁃１００，ＬｏｓＧａｔｏｓＲｅｓｅａｒｃｈ，美国）测量分析不同水体的氘氧稳定同位素比率［２５］：（ɢ）＝（Ｒｓａｍｐｌｅ－Ｒｓｔａｎｄａｒｄ）／Ｒｓｔａｎｄａｒｄˑ１０００（１）式中，δ是样品中氘（或氧）的同位素组成，Ｒｓａｍｐｌｅ和Ｒｓｔａｎｄａｒｄ分别是样品和维也纳国际标准平均海水（ＶＳＭＯＷ）氢（或氧）的重轻同位素丰度之比（２Ｈ／１Ｈ或１８Ｏ／１６Ｏ），的比值，δＤ和δ１８Ｏ的测量精度分别为ʃ１ɢ和ʃ０．１ɢ㊂１．３㊀根系水分来源识别本文首先基于实测土壤水和茎秆水的δＤ和δ１８Ｏ双稳定同位素信息，采用直接对比法来初步判别冬小麦不同生育期的根系吸水深度，即通过对比同一时期土壤水δＤ和δ１８Ｏ同位素剖面及相应的作物茎秆水双稳定同位素值，当土壤水和茎秆水的δＤ值与两种水体的δ１８Ｏ值存在共同唯一交点时，该点对应的深度即可判断为冬小麦的主要根系吸水深度［２，１２］；当存在两个或多个共同交点时，根系吸水深度不唯一；若无共同交点，则难以确定其主要吸水深度［２，１２］㊂同时，采用基于贝叶斯理论的ＭｉｘＳＩＡＲ模型来量化冬小麦水分来源及其贡献比例㊂由于研究区地下水位埋深（１６ｍ）远大于华北平原地下水蒸发的极限水位埋深（４ ８ｍ）［２６⁃２７］，因此本研究中植物水分来源不考虑地下水㊂小麦吸收利用的水分主要来自不同深度的土壤水，是大气降水㊁灌溉水和早期土壤水的混合体㊂４１６６㊀生㊀态㊀学㊀报㊀㊀㊀３８卷㊀根据土壤剖面含水量㊁土壤水同位素组成和根系分布特征，将２ｍ土体划分为０ ２０㊁２０ ７０㊁７０ １５０ｃｍ和１５０ ２００ｃｍ４层进行水源分析㊂由于土壤水样采集深度为１０，２０，３０，５０，７０，９０，１１０，１５０ｃｍ和２００ｃｍ，量化水分来源时４个土层对应的土壤水同位素值分别取该层范围内所有采样深度土壤水同位素值的平均值㊂ＭｉｘＳＩＡＲ模型的输入数据包括源（各层土壤水δＤ和δ１８Ｏ双稳定同位素实测值）数据和混合物（茎秆水）数据，ＭａｒｋｏｖＣｈａｉｎＭｏｎｔｅＣａｒｌｏ（ＭＣＭＣ）运行步长为 ｖｅｒｙｌｏｎｇ ，模型误差选取 ｐｒｏｃｅｓｓ＋ｒｅｓｉｄｕａｌ ，由此估算得到的每个水源相应的中值（５０％分位数）贡献比例即视为该水源对植物水的贡献率㊂２㊀结果２．１㊀不同水体同位素分布特征２０１４和２０１５年冬小麦返青至收获试验期间当地大气降水线（ＬＭＷＬ）分别为：δＤ＝７．３δ１８Ｏ＋３．６（Ｒ２＝０．９７，Ｐ＜０．０１）和δＤ＝６．７δ１８Ｏ＋１．８（Ｒ２＝０．９７，Ｐ＜０．０１）（图１）㊂２０１５季降水过程中蒸发速率较快，其ＬＭＷＬ斜率明显小于２０１４季ＬＭＷＬ的斜率㊂土壤水蒸发作用强烈，其同位素值基本位于ＬＭＷＬ的右下方，δＤ δ１８Ｏ拟合线斜率显著低于ＬＷＭＬ㊂２０１５季土壤水δＤ δ１８Ｏ拟合线的斜率（２．８）低于２０１４季（４．０），土壤水蒸发强度更大㊂由图１可知，不同深度土壤水同位素值差异显著，表层０ ２０ｃｍ土壤水同位素最为富集，且变化幅度最大，表明０ ２０ｃｍ土壤水发生强蒸发作用，同时受降雨灌溉补给影响㊂土壤水同位素值变化幅度随土壤深度的增加而逐渐减小㊂试验期间１５０ ２００ｃｍ深度土壤水同位素值接近灌溉水的同位素值且无显著变化㊂作物茎水的同位素基本位于土壤水δＤ δ１８Ｏ同位素拟合线附近，２０１４季总体落在０ １５０ｃｍ深度土壤水同位素值范围内，而２０１５季则比较富集且主要落在浅层０ ７０ｃｍ土壤水同位素值区间内，由此可推测２０１４和２０１５季冬小麦分别吸收利用０ １５０ｃｍ和０ ７０ｃｍ深度的土壤水㊂图１㊀不同水体δＤ⁃δ１８Ｏ关系和当地大气降水线（ＬＭＷＬ）Ｆｉｇ．１㊀ＴｈｅｒｅｌａｔｉｏｎｓｈｉｐｂｅｔｗｅｅｎδＤａｎｄδ１８Ｏｉｎｄｉｆｆｅｒｅｎｔｗａｔｅｒｓａｎｄｔｈｅｌｏｃａｌｍｅｔｅｏｒｏｌｏｇｉｃａｌｗａｔｅｒｌｉｎｅ（ＬＭＷＬ）ｉｎｔｈｅ２０１４ａｎｄ２０１５ｇｒｏｗｉｎｇｓｅａｓｏｎｓ２．２㊀冬小麦不同生育期根系吸水深度由直接对比法分析可知，２０１４季冬小麦返青⁃拔节㊁拔节⁃抽穗㊁抽穗⁃灌浆和灌浆⁃收获期根系吸水深度分别为０ ２０㊁２０ ７０㊁７０ １５０ｃｍ和７０ １５０ｃｍ（图２），而２０１５季各生育期的吸水深度则为０ ２０㊁２０ ７０㊁０ ２０ｃｍ和２０ ７０ｃｍ（图３）㊂不同处理的根系吸水深度季节变化差异显著㊂２０１４季返青⁃拔节期Ｔ１和Ｔ５处理的根系吸水深度为０ ２０ｃｍ，Ｔ２为０ ２０ｃｍ和２０ ７０ｃｍ，Ｔ３和Ｔ４均为２０ ７０ｃｍ；拔节⁃抽穗期Ｔ１和５１６６１８期㊀㊀㊀杜俊杉㊀等：基于双稳定同位素和ＭｉｘＳＩＡＲ模型的冬小麦根系吸水来源研究㊀６１６６㊀生㊀态㊀学㊀报㊀㊀㊀３８卷㊀Ｔ５处理的根系吸水深度为２０ ７０ｃｍ，Ｔ３处理为０ ２０ｃｍ和２０ ７０ｃｍ，而Ｔ４处理则深达７０ １５０ｃｍ；抽穗⁃灌浆期Ｔ２和Ｔ５处理的根系吸水深度深达７０ １５０ｃｍ，Ｔ３和Ｔ４处理冬小麦则吸收２０ ７０ｃｍ土壤水；灌浆⁃收获期除Ｔ１处理吸水深度回到浅层土壤（０ ２０ｃｍ和２０ ７０ｃｍ）外，其余处理的吸水深度仍位于７０ １５０ｃｍ㊂２０１５季不同处理间冬小麦吸水深度的差异主要体现在拔节⁃抽穗期，其中Ｔ１㊁Ｔ２和Ｔ３吸水深度位于图２㊀２０１４季冬小麦各生育阶段土壤水和茎秆水氘氧同位素分布和根系吸水深度Ｆｉｇ．２㊀δＤａｎｄδ１８Ｏｉｎｓｏｉｌｗａｔｅｒａｎｄｓｔｅｍｗａｔｅｒａｎｄｔｈｅｒｏｏｔｗａｔｅｒｕｐｔａｋｅｄｅｐｔｈａｔｅａｃｈｇｒｏｗｔｈｓｔａｇｅｏｆｗｉｎｔｅｒｗｈｅａｔｉｎｔｈｅ２０１４ｓｅａｓｏｎｆｏｒｔｒｅａｔｍｅｎｔｓＴ１：处理１；Ｔ２：处理２；Ｔ３：处理３；Ｔ４：处理４；Ｔ５：处理５深层土壤（７０ ２００ｃｍ），而Ｔ４和Ｔ５处理则位于浅层（０ ７０ｃｍ）㊂２０１４季拔节⁃抽穗期Ｔ２处理土壤水和茎秆水δＤ值的交叉点（７０ １５０ｃｍ和１５０ ２００ｃｍ）与土壤水和茎秆水δ１８Ｏ值的交叉点（０ ２０ｃｍ和２０ ７０ｃｍ）存在明显差异，Ｔ１处理由δＤ和δ１８Ｏ同位素推断的抽穗⁃灌浆期根系吸水深度（分别为２０ ７０ｃｍ和７０图３㊀２０１５季冬小麦各生育阶段土壤水和茎秆水氘氧同位素分布和根系吸水深度Ｆｉｇ．３㊀δＤａｎｄδ１８Ｏｉｎｓｏｉｌｗａｔｅｒａｎｄｓｔｅｍｗａｔｅｒａｎｄｔｈｅｒｏｏｔｗａｔｅｒｕｐｔａｋｅｄｅｐｔｈａｔｅａｃｈｇｒｏｗｔｈｓｔａｇｅｏｆｗｉｎｔｅｒｗｈｅａｔｉｎｔｈｅ２０１５ｓｅａｓｏｎｆｏｒｔｒｅａｔｍｅｎｔｓ１５０ｃｍ）亦不同，而灌浆⁃收获期Ｔ１㊁Ｔ３和Ｔ４处理又同时存在两个吸水深度，此时，利用直接对比法难以准确判定作物的主要吸水深度㊂２．３㊀冬小麦吸水来源及其贡献比例利用ＭｉｘＳＩＡＲ贝叶斯同位素混合模型定量计算２０１４和２０１５季冬小麦返青⁃收获期间不同处理各深度的土壤水对作物水的贡献比例，并确定各生育阶段冬小麦的主要根系吸水深度㊂结果表明，０ ２０㊁２０ ７０㊁７０ １５０ｃｍ和１５０ ２００ｃｍ深度的土壤水对作物水平均贡献比例分别为３５．６％㊁２７．６％㊁２３．１％和１３．７％㊂但两个生长季冬小麦水分来源差异显著㊂２０１４季冬小麦的主要根系吸水深度随作物的生长发育而逐渐增加，返青⁃７１６６１８期㊀㊀㊀杜俊杉㊀等：基于双稳定同位素和ＭｉｘＳＩＡＲ模型的冬小麦根系吸水来源研究㊀８１６６㊀生㊀态㊀学㊀报㊀㊀㊀３８卷㊀拔节㊁拔节⁃抽穗㊁抽穗⁃灌浆和灌浆⁃收获期的主要吸水深度及其土壤水贡献比例分别为０ ２０ｃｍ（６３．６％）㊁２０ ７０ｃｍ（６７．９％）㊁７０ １５０ｃｍ（５４．４％）和７０ １５０ｃｍ（３９．８％）（图４）㊂图４㊀２０１４和２０１５季冬小麦各生育期土壤水贡献比例（平均值ʃ标准差）Ｆｉｇ．４㊀Ｐｒｏｐｏｒｔｉｏｎｓｏｆｓｏｉｌｗａｔｅｒｃｏｎｔｒｉｂｕｔｉｏｎｓｔｏｗｉｎｔｅｒｗｈｅａｔａｔｅａｃｈｇｒｏｗｔｈｓｔａｇｅｉｎ２０１４ａｎｄ２０１５ｓｅａｓｏｎｓ（ｍｅａｎʃＳＤ）２０１５季冬小麦主要利用浅层（０ ７０ｃｍ）土壤水分，各生育期内主要吸水深度及其土壤水贡献比例分别为０ ２０ｃｍ（７０．４％）㊁７０ １５０ｃｍ（３２．２％）㊁０ ２０ｃｍ（６３．４％）和２０ ７０ｃｍ（５４．９％）（图４）㊂其中，２０１５季０ ２０ｃｍ深度土壤水的贡献比例明显高于２０１４季（１３．９％），但拔节⁃抽穗期受干旱气候（降水比２０１４年减少了２６．９ｍｍ）影响，７０ １５０ｃｍ和１５０ ２００ｃｍ深层土壤水的贡献比例（３２．２％和２３．５％）显著提升㊂不同灌溉施肥条件下冬小麦拔节⁃抽穗期根系吸水来源年季变化差异显著㊂受气候干旱影响，低施肥（１５０ｋｇ／ｈｍ２Ｎ）处理Ｔ３２０１５季主要根系吸水深度大于２０１４季，１５０ ２００ｃｍ土壤水的贡献比例增至３８．９％㊂Ｔ１和Ｔ２处理２０１５季比２０１４季减少灌水（８０ｍｍ）一次，导致水分亏缺加剧，且作物迅速生长㊁耗水速率增大，促使Ｔ１处理１５０ ２００ｃｍ（４７．３％）和Ｔ２处理７０ １５０ｃｍ（６６．６％）的土壤水贡献比例最大㊂当返青⁃拔节期灌水充足（８０ｍｍ）且施肥量ȡ当地施肥量２１０ｋｇ／ｈｍ２Ｎ时，２０１５季拔节⁃抽穗期气候干旱反而促使Ｔ４和Ｔ５处理表层０ ２０ｃｍ根系吸水比例分别增加５１．８％和６３．６％㊂２０１５季抽穗⁃灌浆期复水后，Ｔ１㊁Ｔ２和Ｔ５处理灌浆⁃收获期主要根系吸水深度（２０ ７０ｃｍ）明显小于２０１４季（７０ １５０ｃｍ）㊂３㊀讨论３．１㊀直接对比法和ＭｉｘＳＩＡＲ模型比较采用直接对比法和ＭｉｘＳＩＡＲ模型识别的不同生育阶段冬小麦根系吸水深度结果基本一致㊂但直接对比法仅限于推测单一吸水深度，忽略了多种水源组合对冬小麦水分利用的影响，其判别结果具有不确定性；且当茎秆水与土壤水的δＤ和δ１８Ｏ同位素剖面无共同交点或者存在多个交点时，无法准确判定其主要吸水深度㊂ＭｉｘＳＩＡＲ模型则基于贝叶斯统计理论，结合双稳定同位素剖面分布特征，考虑多水源混合作用，提高了不同深度土壤水对作物水贡献比例计算的准确性㊂例如，本文采用ＭｉｘＳＩＡＲ模型确定的冬小麦返青⁃拔节和拔节⁃抽穗期主要吸水深度０ ２０ｃｍ和２０ ７０ｃｍ，与前人研究中王鹏［２］（分别为０ ２０ｃｍ和４０ ７０ｃｍ）和安江龙等［２１］（分别为０ ２０ｃｍ和０ ４０ｃｍ）的结果基本吻合㊂但后期由于受降雨㊁灌溉和施肥等因素影响，研究结果存在一定差异㊂苑晶晶等［１５］研究表明冬小麦返青期内作物根系主要分布在表层，表层土壤水的贡献比例最高，此后受降雨和灌溉分布影响，不同处理间根系吸水深度差异明显，与本文不同灌溉施肥处理根系吸水深度年季变化差异相一致，从而进一步验证了ＭｉｘＳＩＡＲ模型计算结果的可靠性㊂３．２㊀根系吸水与土壤水分变化关系冬小麦返青至收获期间０ ２００ｃｍ剖面土壤储水量平均减少了１２７．９ｍｍ，其中，０ １５０ｃｍ深度的土壤水消耗量约占９２％㊂由图５可知，１５０ ２００ｃｍ深度土壤含水量变化不大，与主要根系吸水深度在１５０ｃｍ以内（贡献比例８６．３％）一致㊂２０１５季冬小麦根系吸水深度为０ ７０ｃｍ，其０ ７０ｃｍ土壤水消耗量显著高于２０１４季（平均多３９ｍｍ）（图５）㊂２０１５季拔节⁃抽穗期气候干旱加之未补充灌溉，导致该时期土壤储水减少量最大（９８．１ｍｍ），约占全生育期土壤储水总消耗量的６７．４％㊂不同处理间土壤水变化差异显著，Ｔ４处理土壤储水消耗量最多，２０１４和２０１５季分别达１５１．８ｍｍ和１７４．５ｍｍ㊂常规灌溉施肥处理Ｔ５在２０１４生长季内土壤水消耗量最少（８０．３ｍｍ），但２０１５季土壤水消耗量显著增加，总耗水量达１４３．４ｍｍ㊂冬小麦试验期间土壤水的主要补充水源为降水和灌溉水，降水分布和不同处理间灌溉制度的差异直接影响了土壤水分分布的季节变化㊂此外，不同施肥量导致作物养分吸收和生长动态不同，对作物水分利用过程产生了重要影响㊂在华北平原，冬小麦最大根深可达２ｍ，根系主要分布在８０ｃｍ以内，其中０ ４０ｃｍ土层根量和根长密度最大［２８］㊂灌水次数越多表层根量越大［２８］，干旱环境下增施氮肥会刺激表层根系生长［２９］，且土壤水贡献比例与根长密度占比显著线性相关［２３］，因此，２０１５季Ｔ４和Ｔ５处理拔节⁃抽穗期冬小麦主要利用表层（０ ２０ｃｍ）土壤水分㊂低氮处理则促进深层根系发育，Ｔ１和Ｔ３处理深层土壤（１５０ ２００ｃｍ）贡水比例显著增加４３．１％㊂而干旱环境下表层土壤水分严重亏缺时作物深层根系较发达且侧生根增多［３０］，８０ｃｍ以下深层土壤水分消耗比例较大［２８］，从而证明了２０１５季灌水量较少的Ｔ１和Ｔ２处理平均约８５％的作物水来源于深层土壤（７０ ２００ｃｍ）水㊂干旱后复水灌溉使得冬小麦对表层土壤水分利用非常明显［１５］，２０１５季抽穗⁃灌浆期根系吸水深度回到表层０ ２０ｃｍ㊂在今后的研究中，将补充测定不同时期作物根系分布和根系水的同位素组成，从而进一步验证本研究结果㊂４㊀结论１）冬小麦返青后，０ ２０㊁２０ ７０㊁７０ １５０ｃｍ和１５０ ２００ｃｍ深度的土壤水对作物水平均贡献比例分别为３５．６％㊁２７．６％㊁２３．１％和１３．７％㊂返青⁃拔节㊁拔节⁃抽穗㊁抽穗⁃灌浆和灌浆⁃收获期的主要根系吸水深度分别为０ ２０㊁２０ ７０㊁０ ２０ｃｍ和２０ ７０ｃｍ㊂２）受降水灌溉分布影响，２０１４和２０１５年季间冬小麦主要根系吸水深度差异显著㊂２０１４季主要吸水深度从返青⁃拔节期的０ ２０ｃｍ（６３．６％）逐渐增加至抽穗⁃灌浆的７０ １５０ｃｍ（６７．９％），灌浆⁃收获期保持在７０ １５０ｃｍ（５４．４％）；而２０１５季冬小麦则主要利用浅层（０ ７０ｃｍ）土壤水分，其中０ ２０ｃｍ深度土壤水的贡献比例明显高于２０１４季（１３．９％）㊂３）不同处理间根系吸水来源的差异主要发生在作物生长中期，尤其在气候干旱和无补充灌溉条件下的２０１５季拔节⁃抽穗期，返青⁃抽穗期仅灌水２０ｍｍ（Ｔ１和Ｔ２处理）或施肥１０５ｋｇ／ｈｍ２Ｎ（Ｔ３处理）条件下拔节⁃抽穗期深层（７０ ２００ｃｍ）土壤水分利用率平均增加２９％，但当前期充分灌水且大量施肥（ȡ当地施肥量２１０ｋｇ／ｈｍ２Ｎ）时根系吸水深度则位于土壤表层０ ２０ｃｍ（Ｔ４和Ｔ５处理）㊂９１６６１８期㊀㊀㊀杜俊杉㊀等：基于双稳定同位素和ＭｉｘＳＩＡＲ模型的冬小麦根系吸水来源研究㊀图５㊀２０１４和２０１５年冬小麦返青后降水㊁灌溉和土壤含水量分布Ｆｉｇ．５㊀Ｐｒｅｃｉｐｉｔａｔｉｏｎ，ｉｒｒｉｇａｔｉｏｎ，ａｎｄｓｏｉｌｗａｔｅｒｃｏｎｔｅｎｔｄｉｓｔｒｉｂｕｔｉｏｎｏｆｗｉｎｔｅｒｗｈｅａｔａｆｔｅｒｇｒｅｅｎｉｎｇｉｎｔｈｅ２０１４ａｎｄ２０１５ｇｒｏｗｉｎｇｓｅａｓｏｎｓｆｏｒｔｒｅａｔｍｅｎｔｓ０２６６㊀生㊀态㊀学㊀报㊀㊀㊀３８卷㊀㊀㊀４）试验年份冬小麦生长期为枯水季（平均降水量８６．９ｍｍ），作物水分来源与土壤水分分布密切相关，０ １５０ｃｍ深度内土壤水对作物水的贡献比例（８６．３％）与该深度内土壤储水减少量占比（９２％）基本一致㊂该研究表明，ＭｉｘＳＩＡＲ模型可简单有效的定量计算农田根系吸水来源，为合理制定灌水深度和施肥量提供了新的方法㊂参考文献（Ｒｅｆｅｒｅｎｃｅｓ）：［１］㊀ＡｓｂｊｏｒｎｓｅｎＨ，ＭｏｒａＧ，ＨｅｌｍｅｒｓＭＪ．ＶａｒｉａｔｉｏｎｉｎｗａｔｅｒｕｐｔａｋｅｄｙｎａｍｉｃｓａｍｏｎｇｃｏｎｔｒａｓｔｉｎｇａｇｒｉｃｕｌｔｕｒａｌａｎｄｎａｔｉｖｅｐｌａｎｔｃｏｍｍｕｎｉｔｉｅｓｉｎｔｈｅＭｉｄｗｅｓｔｅｒｎＵ．Ｓ．Ａｇｒｉｃｕｌｔｕｒｅ，Ｅｃｏｓｙｓｔｅｍｓ＆Ｅｎｖｉｒｏｎｍｅｎｔ，２００７，１２１（４）：３４３⁃３５６．［２］㊀王鹏．基于氢氧稳定同位素的农田ＳＰＡＣ系统水分运移规律研究：以山西省运城市董村农场为例［Ｄ］．北京：中国科学院研究生院，２０１０．［３］㊀ＺｈａｎｇＹＣ，ＳｈｅｎＹＪ，ＳｕｎＨＹ，ＧａｔｅｓＪＢ．Ｅｖａｐｏｔｒａｎｓｐｉｒａｔｉｏｎａｎｄｉｔｓｐａｒｔｉｔｉｏｎｉｎｇｉｎａｎｉｒｒｉｇａｔｅｄｗｉｎｔｅｒｗｈｅａｔｆｉｅｌｄ：ａｃｏｍｂｉｎｅｄｉｓｏｔｏｐｉｃａｎｄｍｉｃｒｏｍｅｔｅｏｒｏｌｏｇｉｃａｐｐｒｏａｃｈ．ＪｏｕｒｎａｌｏｆＨｙｄｒｏｌｏｇｙ，２０１１，４０８（３／４）：２０３⁃２１１．［４］㊀ＺｉｍｍｅｒｍａｎｎＵ，ＥｈｈａｌｔＤ，ＭｕｎｎｉｃｈＫＯ．Ｓｏｉｌｗａｔｅｒｍｏｖｅｍｅｎｔａｎｄｅｖａｐｏｔｒａｎｓｐｉｒａｔｉｏｎ：ｃｈａｎｇｅｓｉｎｔｈｅｉｓｏｔｏｐｉｃｃｏｍｐｏｓｉｔｉｏｎｏｆｔｈｅｗａｔｅｒ／／ＰｒｏｃｅｅｄｉｎｇｓｏｆＩＡＥＡＳｙｍｐｏｓｉｕｍｏｎＩｓｏｔｏｐｅｓｉｎＨｙｄｒｏｌｏｇｙ．Ｖｉｅｎｎａ：ＩＡＥＡ，１９６７：５６７⁃５８４．［５］㊀ＳｐｒｅｎｇｅｒＭ，ＬｅｉｓｔｅｒｔＨ，ＧｉｍｂｅｌＫ，ＷｅｉｌｅｒＭ．Ｉｌｌｕｍｉｎａｔｉｎｇｈｙｄｒｏｌｏｇｉｃａｌｐｒｏｃｅｓｓｅｓａｔｔｈｅｓｏｉｌ⁃ｖｅｇｅｔａｔｉｏｎ⁃ａｔｍｏｓｐｈｅｒｅｉｎｔｅｒｆａｃｅｗｉｔｈｗａｔｅｒｓｔａｂｌｅｉｓｏｔｏｐｅｓ．ＲｅｖｉｅｗｓｏｆＧｅｏｐｈｙｓｉｃｓ，２０１６，５４（３）：６７４⁃７０４．［６］㊀ＢｒｕｎｅｌＪＰ，ＷａｌｋｅｒＧＲ，Ｋｅｎｎｅｔｔ⁃ＳｍｉｔｈＡＫ．Ｆｉｅｌｄｖａｌｉｄａｔｉｏｎｏｆｉｓｏｔｏｐｉｃｐｒｏｃｅｄｕｒｅｓｆｏｒｄｅｔｅｒｍｉｎｉｎｇｓｏｕｒｃｅｓｏｆｗａｔｅｒｕｓｅｄｂｙｐｌａｎｔｓｉｎａｓｅｍｉ⁃ａｒｉｄｅｎｖｉｒｏｎｍｅｎｔ．ＪｏｕｒｎａｌｏｆＨｙｄｒｏｌｏｇｙ，１９９５，１６７（１／４）：３５１⁃３６８．［７］㊀ＳｎｙｄｅｒＫＡ，ＷｉｌｌｉａｍｓＤＧ．ＷａｔｅｒｓｏｕｒｃｅｓｕｓｅｄｂｙｒｉｐａｒｉａｎｔｒｅｅｓｖａｒｉｅｓａｍｏｎｇｓｔｒｅａｍｔｙｐｅｓｏｎｔｈｅＳａｎＰｅｄｒｏＲｉｖｅｒ，Ａｒｉｚｏｎａ．ＡｇｒｉｃｕｌｔｕｒａｌａｎｄＦｏｒｅｓｔＭｅｔｅｏｒｏｌｏｇｙ，２０００，１０５（１／３）：２２７⁃２４０．［８］㊀ＰｈｉｌｌｉｐｓＤＬ，ＧｒｅｇｇＪＷ．Ｓｏｕｒｃｅｐａｒｔｉｔｉｏｎｉｎｇｕｓｉｎｇｓｔａｂｌｅｉｓｏｔｏｐｅｓ：ｃｏｐｉｎｇｗｉｔｈｔｏｏｍａｎｙｓｏｕｒｃｅｓ．Ｏｅｃｏｌｏｇｉａ，２００３，１３６（２）：２６１⁃２６９．［９］㊀ＭｏｏｒｅＪＷ，ＳｅｍｍｅｎｓＢＸ．Ｉｎｃｏｒｐｏｒａｔｉｎｇｕｎｃｅｒｔａｉｎｔｙａｎｄｐｒｉｏｒｉｎｆｏｒｍａｔｉｏｎｉｎｔｏｓｔａｂｌｅｉｓｏｔｏｐｅｍｉｘｉｎｇｍｏｄｅｌｓ．ＥｃｏｌｏｇｙＬｅｔｔｅｒｓ，２００８，１１（５）：４７０⁃４８０．［１０］㊀ＰａｒｎｅｌｌＡＣ，ＩｎｇｅｒＲ，ＢｅａｒｈｏｐＳ，ＪａｃｋｓｏｎＡＬ．Ｓｏｕｒｃｅｐａｒｔｉｔｉｏｎｉｎｇｕｓｉｎｇｓｔａｂｌｅｉｓｏｔｏｐｅｓ：ｃｏｐｉｎｇｗｉｔｈｔｏｏｍｕｃｈｖａｒｉａｔｉｏｎ．ＰＬｏＳＯｎｅ，２０１０，５（３）：ｅ９６７２．［１１］㊀ＳｔｏｃｋＢＣ，ＳｅｍｍｅｎｓＢＸ．ＭｉｘＳＩＡＲＧＵＩＵｓｅｒＭａｎｕａｌ，ｖｅｒｓｉｏｎ１．０．ｈｔｔｐ：／／ｃｏｎｓｅｒｖｅｒ．ｉｕｇｏ⁃ｃａｆｅ．ｏｒｇ／ｕｓｅｒ／ｂｒｉｃｅ．ｓｅｍｍｅｎｓ／ＭｉｘＳＩＡＲ．［１２］㊀ＹａｎｇＢ，ＷｅｎＸＦ，ＳｕｎＸＭ．ＳｅａｓｏｎａｌｖａｒｉａｔｉｏｎｓｉｎｄｅｐｔｈｏｆｗａｔｅｒｕｐｔａｋｅｆｏｒａｓｕｂｔｒｏｐｉｃａｌｃｏｎｉｆｅｒｏｕｓｐｌａｎｔａｔｉｏｎｓｕｂｊｅｃｔｅｄｔｏｄｒｏｕｇｈｔｉｎａｎＥａｓｔＡｓｉａｎｍｏｎｓｏｏｎｒｅｇｉｏｎ．ＡｇｒｉｃｕｌｔｕｒａｌａｎｄＦｏｒｅｓｔＭｅｔｅｏｒｏｌｏｇｙ，２０１５，２０１（２）：２１８⁃２２８．［１３］㊀刘树宝，陈亚宁，陈亚鹏，邓海军，方功焕．基于稳定同位素技术的黑河下游不同林龄胡杨的吸水深度研究．生态学报，２０１６，３６（３）：７２９⁃７３９．［１４］㊀ＭｃＣｏｌｅＡＡ，ＳｔｅｒｎＬＡ．ＳｅａｓｏｎａｌｗａｔｅｒｕｓｅｐａｔｔｅｒｎｓｏｆＪｕｎｉｐｅｒｕｓａｓｈｅｉｏｎｔｈｅＥｄｗａｒｄｓＰｌａｔｅａｕ，Ｔｅｘａｓ，ｂａｓｅｄｏｎｓｔａｂｌｅｉｓｏｔｏｐｅｓｉｎｗａｔｅｒ．ＪｏｕｒｎａｌｏｆＨｙｄｒｏｌｏｇｙ，２００７，３４２（３／４）：２３８⁃２４８．［１５］㊀苑晶晶，袁国富，罗毅，孙晓敏，张娜．利用δ１８Ｏ信息分析冬小麦对浅埋深地下水的利用．自然资源学报，２００９，２４（２）：３６０⁃３６８．［１６］㊀刘自强，余新晓，娄源海，李瀚之，贾国栋，路伟伟．北京山区侧柏水分利用策略．生态学报，２０１７，３７（１１）：３６９７⁃３７０５．［１７］㊀ＺｈａｎｇＣＺ，ＺｈａｎｇＪＢ，ＺｈａｏＢＺ，ＺｈｕＡＮ，ＺｈａｎｇＨ，ＨｕａｎｇＰ，ＬｉＸＰ．Ｃｏｕｐｌｉｎｇａｔｗｏ⁃ｔｉｐｌｉｎｅａｒｍｉｘｉｎｇｍｏｄｅｌｗｉｔｈａδＤ⁃δ１８Ｏｐｌｏｔｔｏｄｅｔｅｒｍｉｎｅｗａｔｅｒｓｏｕｒｃｅｓｃｏｎｓｕｍｅｄｂｙｍａｉｚｅｄｕｒｉｎｇｄｉｆｆｅｒｅｎｔｇｒｏｗｔｈｓｔａｇｅｓ．ＦｉｅｌｄＣｒｏｐｓＲｅｓｅａｒｃｈ，２０１１，１２３（３）：１９６⁃２０５．［１８］㊀ＷｕＹＪ，ＤｕＴＳ，ＬｉＦＳ，ＬｉＳＥ，ＤｉｎｇＲＳ，ＴｏｎｇＬ．Ｑｕａｎｔｉｆｉｃａｔｉｏｎｏｆｍａｉｚｅｗａｔｅｒｕｐｔａｋｅｆｒｏｍｄｉｆｆｅｒｅｎｔｌａｙｅｒｓａｎｄｒｏｏｔｚｏｎｅｓｕｎｄｅｒａｌｔｅｒｎａｔｅｆｕｒｒｏｗｉｒｒｉｇａｔｉｏｎｕｓｉｎｇｓｔａｂｌｅｏｘｙｇｅｎｉｓｏｔｏｐｅ．ＡｇｒｉｃｕｌｔｕｒａｌＷａｔｅｒＭａｎａｇｅｍｅｎｔ，２０１６，１６８：３５⁃４４．［１９］㊀张景文，陈报章．基于同位素分析研究山东禹城夏玉米水分来源．水土保持学报，２０１７，３１（４）：９９⁃１０４．［２０］㊀李惠，梁杏，刘延锋，刘亚磊，鲜阳．基于氢氧稳定同位素识别干旱区棉花水分利用来源．地球科学，２０１７，４２（５）：８４３⁃８５２．［２１］㊀安江龙，马娟娟，张亚雄，李蕊，孙西欢，郭向红．两种模型分析方法下冬小麦根系吸水深度的对比研究．灌溉排水学报，２０１７，３６（１１）：２５⁃２８．［２２］㊀ＹａｎｇＢ，ＷｅｎＸＦ，ＳｕｎＸＭ．ＩｒｒｉｇａｔｉｏｎｄｅｐｔｈｆａｒｅｘｃｅｅｄｓｗａｔｅｒｕｐｔａｋｅｄｅｐｔｈｉｎａｎｏａｓｉｓｃｒｏｐｌａｎｄｉｎｔｈｅｍｉｄｄｌｅｒｅａｃｈｅｓｏｆＨｅｉｈｅＲｉｖｅｒＢａｓｉｎ．ＳｃｉｅｎｔｉｆｉｃＲｅｐｏｒｔｓ，２０１５，５：１５２０６．［２３］㊀ＭａＹ，ＳｏｎｇＸＦ．Ｕｓｉｎｇｓｔａｂｌｅｉｓｏｔｏｐｅｓｔｏｄｅｔｅｒｍｉｎｅｓｅａｓｏｎａｌｖａｒｉａｔｉｏｎｓｉｎｗａｔｅｒｕｐｔａｋｅｏｆｓｕｍｍｅｒｍａｉｚｅｕｎｄｅｒｄｉｆｆｅｒｅｎｔｆｅｒｔｉｌｉｚａｔｉｏｎｔｒｅａｔｍｅｎｔｓ．１２６６１８期㊀㊀㊀杜俊杉㊀等：基于双稳定同位素和ＭｉｘＳＩＡＲ模型的冬小麦根系吸水来源研究㊀２２６６㊀生㊀态㊀学㊀报㊀㊀㊀３８卷㊀ＳｃｉｅｎｃｅｏｆｔｈｅＴｏｔａｌＥｎｖｉｒｏｎｍｅｎｔ，２０１６，５５０：４７１⁃４８３．［２４］㊀ＣａｉＪＢ，ＬｉｕＹ，ＸｕＤ，ＰａｒｅｄｅｓＰ，ＰｅｒｅｉｒａＬＳ．Ｓｉｍｕｌａｔｉｏｎｏｆｔｈｅｓｏｉｌｗａｔｅｒｂａｌａｎｃｅｏｆｗｈｅａｔｕｓｉｎｇｄａｉｌｙｗｅａｔｈｅｒｆｏｒｅｃａｓｔｍｅｓｓａｇｅｓｔｏｅｓｔｉｍａｔｅｔｈｅｒｅｆｅｒｅｎｃｅｅｖａｐｏｔｒａｎｓｐｉｒａｔｉｏｎ．ＨｙｄｒｏｌｏｇｙａｎｄＥａｒｔｈＳｙｓｔｅｍＳｃｉｅｎｃｅｓ，２００９，１３（７）：１０４５⁃１０５９．［２５］㊀ＷｅｓｔＡＧ，ＰａｔｒｉｃｋｓｏｎＳＪ，ＥｈｌｅｒｉｎｇｅｒＪＲ．Ｗａｔｅｒｅｘｔｒａｃｔｉｏｎｔｉｍｅｓｆｏｒｐｌａｎｔａｎｄｓｏｉｌｍａｔｅｒｉａｌｓｕｓｅｄｉｎｓｔａｂｌｅｉｓｏｔｏｐｅａｎａｌｙｓｉｓ．ＲａｐｉｄＣｏｍｍｕｎｉｃａｔｉｏｎｓｉｎＭａｓｓＳｐｅｃｔｒｏｍｅｔｒｙ，２００６，２０（８）：１３１７⁃１３２１．［２６］㊀雷志栋，杨诗秀，倪广恒，薛迎洲．地下水位埋深类型与土壤水分动态特征．水利学报，１９９２，（２）：１⁃６．［２７］㊀ＬｕｏＹ，ＳｏｐｈｏｃｌｅｏｕｓＭ．Ｓｅａｓｏｎａｌｇｒｏｕｎｄｗａｔｅｒｃｏｎｔｒｉｂｕｔｉｏｎｔｏｃｒｏｐ⁃ｗａｔｅｒｕｓｅａｓｓｅｓｓｅｄｗｉｔｈｌｙｓｉｍｅｔｅｒｏｂｓｅｒｖａｔｉｏｎｓａｎｄｍｏｄｅｌｓｉｍｕｌａｔｉｏｎｓ．ＪｏｕｒｎａｌｏｆＨｙｄｒｏｌｏｇｙ，２０１０，３８９（３／４）：３２５⁃３３５．［２８］㊀ＺｈａｎｇＸＹ，ＰｅｉＤ，ＣｈｅｎＳＹ．ＲｏｏｔｇｒｏｗｔｈａｎｄｓｏｉｌｗａｔｅｒｕｔｉｌｉｚａｔｉｏｎｏｆｗｉｎｔｅｒｗｈｅａｔｉｎｔｈｅＮｏｒｔｈＣｈｉｎａＰｌａｉｎ．ＨｙｄｒｏｌｏｇｉｃａｌＰｒｏｃｅｓｓｅｓ，２００４，１８（１２）：２２７５⁃２２８７．［２９］㊀ＣａｒｖａｌｈｏＰ，ＦｏｕｌｋｅｓＭＪ．Ｒｏｏｔｓａｎｄｕｐｔａｋｅｏｆｗａｔｅｒａｎｄｎｕｔｒｉｅｎｔｓ／／ＣｈｒｉｓｔｏｕＰ，ＳａｖｉｎＲ，Ｃｏｓｔａ⁃ＰｉｅｒｃｅＢＡ，ＭｉｓｚｔａｌＩ，ＷｈｉｔｅｌａｗＣＢＡ，ｅｄｓ．ＳｕｓｔａｉｎａｂｌｅＦｏｏｄＰｒｏｄｕｃｔｉｏｎ．ＮｅｗＹｏｒｋ：Ｓｐｒｉｎｇｅｒ，２０１３：１３９０⁃１４０４．［３０］㊀ＪｈａＳＫ，ＧａｏＹ，ＬｉｕＨ，ＨｕａｎｇＺＤ，ＷａｎｇＧＳ，ＬｉａｎｇＹＰ，ＤｕａｎＡＷ．ＲｏｏｔｄｅｖｅｌｏｐｍｅｎｔａｎｄｗａｔｅｒｕｐｔａｋｅｉｎｗｉｎｔｅｒｗｈｅａｔｕｎｄｅｒｄｉｆｆｅｒｅｎｔｉｒｒｉｇａｔｉｏｎｍｅｔｈｏｄｓａｎｄｓｃｈｅｄｕｌｉｎｇｆｏｒＮｏｒｔｈＣｈｉｎａ．ＡｇｒｉｃｕｌｔｕｒａｌＷａｔｅｒＭａｎａｇｅｍｅｎｔ，２０１７，１８２：１３９⁃１５０．。

第五章沉淀反应—■教学基本要求1、沉淀溶解平衡掌握溶度积常数与溶解度的相互换算。

2、溶度积规则及应用掌握用溶度积规则判断溶液中沉淀的产生和溶解；掌握同离子效应的计算，了解盐效应；掌握计算难溶氢氧化物、硫化物开始沉淀、沉淀完全时c(OH—)、pH值；熟悉沉淀溶解的方法；掌握通过计算判断分步沉淀的丿II页序及第二种离子开始沉淀时，第一种离子是否沉淀完全；掌握沉淀转化反应的平衡常二三§5.1沉淀溶解平衡§5・1・：1溶度积常数将溶解度大于0.1g/100gH20的物质称为易溶电解质，梅溶解度在0.01 0.1g/100gH20的物质称为微溶电解质,将溶解度小于0.01g/100gH20的物质称为难溶电解质。

BaSO4⑸在饱和溶液中存在下列平衡：BaSO4(s) o Ba2*(aq)+SO42-(aq);则:Ksp° = Ba2* SO42- 具中，c—l.Omol/dmq不写入表达式中。

Up称为溶度积常数…即温度一走时,难溶电解质溶在水溶液中的部分『全部离解为离子时离子的浓度的乘积是一常数，简称溶度积。

推广到一般式,如一反应为：AmBn(s) <=> mA n+(aq)+nB^aq)则:K%(AmBn)=冲『.［夕叮即:指定反应式中的离子，以离子的化学计星系数为指数的幕的相对浓度的乘积是一常数。

• K°sp同样是温度的函数，但K°sp受温度影响不大，当温度变化不大时，可采用常温下的资料。

•溶度积的大小反映了难溶电解质溶解能力的大小。

对于同种类型基本不水解难溶强电解质, 溶度积越大,溶解度也越大；对于不同类型难溶电解质，就不能用Kip大小来比较溶解能力的大小,必须把溶度积换算成溶解度。

例如：K e sp S(mol/dm3)AB AgCI 1.8x10“° 1.3x10”AgBr 5.0x10"7.1x10-7A2B Ag2CrO4l.lxlO"12 6.5 xlO^5§5.1.2溶度积常数和溶解度的相互换算难溶化合物的溶解度S和Kip都是表示难溶化合物溶解能力大小的物理墨因此它们之间存在着相互依赖的关系，是可以进行换算的，可以从S求K：p也可以从Up求S。

[模拟] 电厂水处理值班员基础理论知识模拟16简答题第1题：分析天平使用时应做哪些检查工作? ____参考答案：(1)稳定性。

指针左右摆是否灵活有规律。

(2)准确性。

两个质量相等的砝码放在两盘中，天平平衡，两砝码调换后也应平衡，如超过9个分度应检修。

(3)灵敏性。

在天平盘中，放一校对的10mg砝码，开启天平标尺，应在0.9～1.0mg 之内。

(4)示值不变性。

同一物体称量数次，其值误差应不大于1个分度，否则应检修。

第2题：闪蒸装置为防止结垢，常采取什么措施? ____参考答案：闪蒸装置的工作温度较低，它的汽化过程又不在加热面上进行，而且管内含盐水可以维持适当的流速，所以结垢轻微。

为防止含盐水在凝汽器或加热器的传热面上结垢，要保持各凝汽器含盐水侧的压力高于其最高温度所对应的饱和蒸汽压力，在含盐水进入第一级闪蒸室压力降低时，碳酸氢盐分解产生的CO2需要进入一个初级闪蒸室，从这里被抽去。

在闪蒸室中，因没有受热面，故所形成的沉淀物不会变成水垢。

第3题：如何防止反渗透膜结垢? ____参考答案：(1)做好原水的预处理工作，特别应注意污染指数是否合格，同时还应进行杀菌，防止微生物在容器内滋生。

(2)在反渗透设备运行中，要维持合适的操作压力，一般情况下，增加压力会使产水量增大，但过大又会使膜压实。

(3)在反渗透设备运行中，应保持盐水侧的紊流状态，减轻膜表面溶液的浓差极化，避免某些难溶盐在膜表面析出。

(4)在反渗透设备停运时，短期停运应进行加药冲洗，长期停运应加甲醛保护。

(5)当反渗透设备产水量明显减少，表明膜结垢或污染，应进行化学清洗。

第4题：阴阳树脂混杂时，如何将它们分开? ____参考答案：可以利用阴阳树脂密度不同，借自下而上水流分离的方法将它们分开。

另一种方法是将混杂树脂浸泡在饱和食盐水或16%左右的NaOH溶液中，阴树脂就会浮起来，阳树脂则不然。

如果两种树脂密度差很小，则可先将树脂转型，然后再进行分离，因为树脂的类型不同，其密度也发生变化。

9.叙述大气中NO转化为NO2的各种途径。

①NO + O3NO2 + O2②HO + RH R + H2OR + O2 RO2NO + RO2 NO2 + RORO + O2R`CHO + HO2（R`比R少一个C原子）NO + HO2NO2 + HO1．大气中有哪些重要污染物？说明其主要来源和消除途径。

环境中的大气污染物种类很多，若按物理状态可分为气态污染物和颗粒物两大类；若按形成过程则可分为一次污染物和二次污染物。

按照化学组成还可以分为含硫化合物、含氮化合物、含碳化合物和含卤素化合物。

主要按照化学组成讨论大气中的气态污染物主要来源和消除途径如下：（1）含硫化合物大气中的含硫化合物主要包括：氧硫化碳（COS）、二硫化碳（CS2）、二甲基硫（CH3）2S、硫化氢（H2S）、二氧化硫（SO2）、三氧化硫（SO3）、硫酸（H2SO4）、亚硫酸盐（MSO3）和硫酸盐（MSO4）等。

大气中的SO2（就大城市及其周围地区来说）主要来源于含硫燃料的燃烧。

大气中的SO2约有50%会转化形成H2SO4或SO42-，另外50%可以通过干、湿沉降从大气中消除。

H2S主要来自动植物机体的腐烂，即主要由植物机体中的硫酸盐经微生物的厌氧活动还原产生。

大气中H2S主要的去除反应为：HO + H2S → H2O + SH。

（2）含氮化合物大气中存在的含量比较高的氮的氧化物主要包括氧化亚氮（N2O）、一氧化氮（NO）和二氧化氮（NO2）。

主要讨论一氧化氮（NO）和二氧化氮（NO2），用通式NO x表示。

NO和NO2是大气中主要的含氮污染物，它们的人为来源主要是燃料的燃烧。

大气中的NO x最终将转化为硝酸和硝酸盐微粒经湿沉降和干沉降从大气中去除。

其中湿沉降是最主要的消除方式。

（3）含碳化合物大气中含碳化合物主要包括：一氧化碳（CO）、二氧化碳（CO2）以及有机的碳氢化合物（HC）和含氧烃类，如醛、酮、酸等。

CO的天然来源主要包括甲烷的转化、海水中CO的挥发、植物的排放以及森林火灾和农业废弃物焚烧，其中以甲烷的转化最为重要。