City of Pomona v. SQM N.Am. Corp., 750 F.3d 1036 (9th Cir. 2014), pet’n for cert. pending:

After excessive levels of the chemical perchlorate were found in a city's water system, the city undertook to investigate the source of that contamination and remediate. Using a methodology known as "stable isotope analysis," a scientist hired by the city determined that the most likely dominant source of the perchlorate found in the city's groundwater was sodium nitrate that had been used as fertilizer. The sodium nitrate had been imported in large quantities from Chile several decades earlier and had been used as fertilizer over a substantial period of time. The city sued the company that imported the sodium nitrate into the United States. Before trial, the district court held an evidentiary hearing under Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579, 113 S. Ct. 2786, 125 L. Ed. 2d 469 (1993), and excluded the city's expert witness. The parties then entered into a conditional stipulated dismissal to facilitate the appeal of the district court's evidentiary ruling, among other issues. Because the district court abused its discretion by not allowing a jury to resolve contested but otherwise admissible expert testimony, we  [**4] reverse the district court's order of exclusion, affirm the district court's denial of the defendant's motion for summary judgment on other issues, and remand for trial.

 

BACKGROUND

 

The City of Pomona, California ("Pomona"), administers a public water system. Pomona receives its water from the Chino Basin aquifer using a set of 14 wells that connect to Pomona's groundwater treatment facility. In 2007, the Chino Basin aquifer was found to have levels of the chemical perchlorate in excess of the Maximum Contaminant Level ("MCL") of six parts per billion ("ppb") permitted by the California Department of Public Health ("CDPH").

CDPH regulates contaminants in drinking water through several standards, including MCLs and Action Levels. MCLs are legally enforceable numerical standards, statutorily defined as "the maximum permissible level of a contaminant in water." Cal. Health & Safety Code § 116275(f). CDPH has the power to suspend or revoke a municipality's water system operating permit for failure to comply with an MCL. Id. § 116625(a). Action Levels (known as "Notification Levels" after 2004) are non-regulatory advisory levels for contaminants. Id. § 116455(c)(3). The only action required  [**5] when a contaminant exceeds an Action Level, but remains below an MCL or when no MCL has been set, is notification to CDPH. Id. § 116455(a)(2).

In 1999, the CDPH set the perchlorate Action Level at 18 ppb. At this time, consistent with its responsibility under California law, Pomona began monitoring perchlorate levels in its groundwater and reporting these levels to the CDPH. In 2002, the CDPH reduced the perchlorate Action Level to four ppb. Pomona continued to monitor perchlorate levels. In 2007, CDPH established a perchlorate MCL of six ppb. In response to the MCL, Pomona immediately took steps towards compliance, including shutting off wells, purchasing water from other sources, and blending well water with non-well water to reduce the levels of perchlorate. Pomona also began shifting its existing nitrate removal processes to perchlorate removal and hired an engineer to identify a long-term solution for compliance with the MCL.

On October 15, 2010, Pomona filed this lawsuit against SQM North America Corporation ("SQMNA") to recover the cost of investigating and remediating perchlorate contamination in the groundwater in and around Pomona, California. Pomona alleges that SQMNA's importation  [**6] of natural sodium nitrate from the Atacama Desert in Chile for use as a fertilizer was the primary source of Pomona's perchlorate contamination.

 [*1042]  On October 31, 2011, SQMNA moved for summary judgment on two grounds. First, SQMNA argued that Pomona had not suffered a compensable injury under strict products liability law based on California's "economic loss rule." Second, SQMNA argued that even if Pomona had suffered a compensable injury, Pomona's claim was barred by the applicable three-year statute of limitations. The district court denied SQMNA summary judgment on both arguments. The case then proceeded toward trial.

On January 6, 2012, the district court held a Daubert hearing to consider SQMNA's pretrial motion in limine to exclude the testimony of Dr. Neil Sturchio, Pomona's expert witness on causation. Dr. Sturchio is the director of the Environmental Isotope Geochemistry Laboratory at the University of Illinois at Chicago. Dr. Sturchio began working on Pomona's perchlorate case in April 2011, using a methodology known as "stable isotope analysis."¹

1   An atom is a basic unit of matter that consists of a central or core nucleus surrounded by a cloud of negatively charged electrons.  [**7] Inside the nucleus are positively charged protons and, typically, electrically neutral neutrons. An isotope is a form of a chemical element that has the same number of protons in the nucleus (i.e., the same atomic number) as that element but a different number of neutrons in the nucleus (i.e., a different atomic weight). Isotopes that are not subject to nuclear decay are known as "stable isotopes," whereas isotopes that are subject to nuclear decay are known as "radioactive isotopes." Isotope analysis is the study of the nucleus of an atom. Stable isotope analysis is based on the proposition that stable isotopes of a given chemical element (e.g., perchlorate) can have distinct isotopic compositions that may indicate the origin or source of a molecule containing that element.

Acting under the direction of Dr. Sturchio, Wildermuth Environmental, Inc. ("Wildermuth") collected well water samples from Pomona using methods based on the Guidance Manual for Forensic Analysis of Perchlorate in Groundwater using Chlorine and Oxygen Isotopic Analyses ("Guidance Manual"). Wildermuth shipped those samples to Dr. Sturchio with blind labels. Dr. Sturchio analyzed the isotopic composition of the perchlorate  [**8] in Pomona's groundwater using stable isotope analysis and compared the resulting information with a reference database of known perchlorate sources.

Dr. Sturchio used a four-step methodology with multiple sub-parts. Dr. Sturchio disclosed this methodology in his expert report filed in this litigation. It was also published in 2011 in the Guidance Manual, which was commissioned by the Environmental Security Technology Certification Program ("ESTCP") of the United States Department of Defense. The four steps described in the Guidance Manual are: (1) collection of groundwater samples; (2) extraction and purification; (3) oxygen and chlorine isotopic analyses on the purified samples; and (4) determination of probable sources by comparing the resulting isotope data to a reference database. Before the publication of the Guidance Manual, peer-reviewed articles provided abbreviated descriptions of the fundamental methods used for stable isotope analysis by Dr. Sturchio and his colleagues.

Based on this analysis, Dr. Sturchio opined that the dominant source of perchlorate in the Pomona groundwater is from the Atacama Desert in Chile and that the samples also contained minor amounts of perchlorate  [**9] from other non-Atacama sources, including synthetic or indigenous natural sources. Based largely upon Dr. Sturchio's findings, Pomona argued that the perchlorate found in its groundwater had the same distinctive isotopic composition as the perchlorate imported into  [*1043]  southern California from Chile by SQMNA between 1927 and the 1950s.

SQMNA moved to exclude Dr. Sturchio's opinions, arguing that "stable isotope analysis" failed to satisfy Daubert and was insufficiently reliable to be received in evidence under Rule 702 of the Federal Rules of Evidence. After an evidentiary hearing, the district court granted SQMNA's motion in limine to  [**10] exclude Dr. Sturchio's testimony. The district court excluded Dr. Sturchio's opinions as unreliable on the grounds that: (1) the opinions were subject to future methodological revisions and not yet certified; (2) the procedures he used had not yet been tested and were not subject to retesting; and (3) the reference database used by Dr. Sturchio was too small. Shortly thereafter, Pomona and SQMNA stipulated to a conditional dismissal with prejudice in order to facilitate review of the district court's order excluding Dr. Sturchio's testimony, among other rulings.²

2   Pomona also argued that the district court abused its discretion by failing expressly to apply the factors considered in Daubert v. Merrell Dow Pharmaceuticals, Inc., 43 F.3d 1311 (9th Cir. 1995) ("Daubert II"). "[W]hether Daubert's specific factors are, or are not, reasonable measures of reliability in a particular case is a matter that the law grants the trial judge broad latitude to determine." Kumho Tire Co., Ltd. v. Carmichael, 526 U.S. 137, 153, 119 S. Ct. 1167, 143 L. Ed. 2d 238 (1999). The district court cited Daubert as the controlling rule of law in evaluating SQMNA's motion to exclude Dr. Sturchio's testimony. The district court did not abuse its  [**11] discretion by not explicitly reciting the factors analyzed in Daubert II. See United States v. Preston, 706 F.3d 1106, 1118 (9th Cir. 2013).

STANDARDS OF REVIEW

We review evidentiary rulings for abuse of discretion and reverse if the exercise of discretion is both erroneous and prejudicial. Nevada Dept. of Corr. v. Greene, 648 F.3d 1014, 1018 (9th Cir. 2011) (citation omitted). We review underlying factual determinations for clear error. United States v. Lukashov, 694 F.3d 1107, 1114 (9th Cir. 2012). We review a district court's order granting or denying summary judgment de novo. Ford v. City of Yakima, 706 F.3d 1188, 1192 (9th Cir. 2013).

DISCUSSION

A. Exclusion of Pomona's Expert Witness Dr. Sturchio

1. Legal Standards

Rule 702 of the Federal Rules of Evidence provides that expert opinion evidence is admissible if: (1) the witness is sufficiently qualified as an expert by knowledge, skill, experience, training, or education; (2) the scientific, technical, or other specialized knowledge will help the trier of fact to understand the evidence or to determine a fact in issue; (3) the testimony is based on sufficient facts or data; (4) the testimony is the product of reliable principles and  [**12] methods; and (5) the expert has reliably applied the relevant principles and methods to the facts of the case. Fed. R. Evid. 702.

Under Daubert and its progeny, including Daubert II, a district court's inquiry into admissibility is a flexible one. Alaska Rent-A-Car, Inc. v. Avis Budget Grp., Inc., 738 F.3d 960, 969 (9th Cir. 2013). In evaluating proffered expert testimony, the trial court is "a gatekeeper, not a fact finder." Primiano v. Cook, 598 F.3d 558, 565 (9th Cir. 2010) (citation and quotation marks omitted).

"[T]he trial court must assure that the expert testimony 'both rests on a reliable foundation and is relevant to the task at hand.'" Id. at 564 (quoting Daubert, [*1044]  509 U.S. at 597). "Expert opinion testimony is relevant if the knowledge underlying it has a valid connection to the pertinent inquiry. And it is reliable if the knowledge underlying it has a reliable basis in the knowledge and experience of the relevant discipline." Id. at 565 (citation and internal quotation marks omitted). "Shaky but admissible evidence is to be attacked by cross examination, contrary evidence, and attention to the burden of proof, not exclusion." Id. at 564 (citation omitted). The judge is "supposed  [**13] to screen the jury from unreliable nonsense opinions, but not exclude opinions merely because they are impeachable." Alaska Rent-A-Car, 738 F.3d at 969. Simply put, "[t]he district court is not tasked with deciding whether the expert is right or wrong, just whether his testimony has substance such that it would be helpful to a jury." Id. at 969-70.

The test of reliability is flexible. Estate of Barabin v. AstenJohnson, Inc., 740 F.3d 457, 463 (9th Cir. 2014) (en banc). The court must assess the expert's reasoning or methodology, using as appropriate criteria such as testability, publication in peer-reviewed literature, known or potential error rate, and general acceptance. Id.; see also Primiano, 598 F.3d at 564. But these factors are "meant to be helpful, not definitive, and the trial court has discretion to decide how to test an expert's reliability as well as whether the testimony is reliable, based on the particular circumstances of the particular case." Primiano, 598 F.3d at 564 (citations and quotation marks omitted); see also Barabin, 740 F.3d at 463. The test "is not the correctness of the expert's conclusions but the soundness of his methodology," and when an expert meets the  [**14] threshold established by Rule 702, the expert may testify and the fact finder decides how much weight to give that testimony. Primiano, 598 F.3d at 564-65. Challenges that go to the weight of the evidence are within the province of a fact finder, not a trial court judge. A district court should not make credibility determinations that are reserved for the jury.

2. Methodology and certification

The district court concluded that Dr. Sturchio's procedures are not reliable because they are not generally accepted in the scientific community. The court gave two reasons: (1) the Quality Assurance/Quality Control ("QA/QC") parameters were still being refined; and (2) the Environmental Protection Agency ("EPA") has not yet certified stable isotope analysis for organic or inorganic compounds. These reasons are insufficient to exclude Dr. Sturchio's testimony.

First, scientific methods that are subject to "further testing and refinement" may be generally accepted and sufficiently reliable. There are "no certainties in science." Daubert, 509 U.S. at 590. For scientific evidence to be admissible, the proponent must show the assertion is "derived by [a] scientific method." Id. Opinion based on "unsubstantiated  [**15] and undocumented information is the antithesis of . . . scientifically reliable expert opinion." Cabrera v. Cordis Corp., 134 F.3d 1418, 1423 (9th Cir. 1998). The existence of ongoing research, however, does not necessarily invalidate the reliability of expert testimony. See Metabolife Int'l, Inc. v. Wornick, 264 F.3d 832, 843 (9th Cir. 2001) (holding that it was "plain error to hold that the Columbia study was not finished--while the overall project was ongoing, all of the relevant data had been gathered in final form, and Metabolife presented an expert interpretation of that data"). For example, during the "raging controversy" surrounding the new technique of DNA testing, the Ninth Circuit rejected the argument that "the FBI's DNA testing and statistical procedures  [*1045]  may warrant review and revision" as an adequate reason to exclude expert testimony. United States v. Chischilly, 30 F.3d 1144, 1152-53 (9th Cir. 1994).

The controlling standards published in the Guidance Manual are subject to further evolution. A "disagreement over, not an absence of, controlling standards" is not a basis to exclude expert testimony. Chischilly, 30 F.3d at 1154. The methods described in the Guidance Manual  [**16] are the product of 12 peer-reviewed publications on stable isotope analysis of perchlorate. The Guidance Manual is a product of inter-laboratory collaboration that began before the initiation of this litigation. Further, all the methods that Dr. Sturchio used were fully disclosed in his expert report from October 2011. There is no record evidence that Dr. Sturchio's opinion is the product of a hasty, incomplete effort.³

3   SQMNA argues that Dr. Sturchio's analysis is incomplete and was previously excluded by another court. In 2003, the Northern District of Illinois excluded the expert testimony of Dr. Sturchio in a matter that, at best, is tangentially related to the analysis he completed for Pomona. Mejdrech v. Lockformer Co., No. 01 C 6107, 2003 U.S. Dist. LEXIS 15587, 2003 WL 22078388, at *1 (N.D. Ill., Sept. 5, 2003). Although both Mejdrech and this case involve the science of stable isotope analysis, they are factually distinct. In Mejdrech, Dr. Sturchio testified about chlorine isotope ratios between volatile organic compounds taken from the plaintiff's locations and the trichloroethylene (TCE) found on the defendant's property. 2003 U.S. Dist. LEXIS 15587, [WL] at *1. The district court found Dr. Sturchio's opinion to be unreliable due  [**17] to a risk of sample contamination because he departed from peer-reviewed methodologies, because the chlorine isotopes that Dr. Sturchio purported to measure could not be measured on a compound-specific basis (such that he could identify or source specific TCE), and based on the allegation that Dr. Sturchio had failed to address unfavorable results in his expert report. 2003 U.S. Dist. LEXIS 15587, [WL] at *2-3. Dr. Sturchio is using different methodologies in this case, and SQMNA raises unique challenges that are not analogous to the facts of Mejdrech. Thus, the Mejdrech decision has little or no bearing on the analysis here.

Second, the district court noted that because "there are no USEPA-certified methods for CSIA of organic or inorganic compounds," Dr. Sturchio's methods were not reliable. Pomona, however, may satisfy its burden of establishing that the evidence is scientifically valid by "pointing to some objective source to show that the conclusions are based on 'scientific method, as it is practiced by (at least) a recognized minority of scientists in the[] field.'" Southland Sod Farms v. Stover Seed Co., 108 F.3d 1134, 1141 (9th Cir. 1997) (quoting Daubert II, 43 F.3d at 1318-19 (alteration in original)).  [**18] Thus, EPA certification of the isotopic analysis of perchlorate is not a precondition to admissibility.

Dr. Sturchio and two other laboratories compiled the Guidance Manual, which shows that the methods Dr. Sturchio employed were reviewed by other laboratories and subject to inter-laboratory calibration. In particular, Dr. Sturchio has collaborated on the methodology used in this case with Dr. J.K. Böhlke, who is among the world's leading authorities on the measuring and reporting of isotope ratios. This demonstrates that Dr. Sturchio's method is "practiced by (at least) a recognized minority of scientists in the[] field." Id. at 1141. SQMNA attempts to discredit Dr. Sturchio's perchlorate techniques by quoting from an EPA manual on the stable isotope analysis of organic compounds at hazardous waste sites. The statements in the EPA manual relating to hazardous waste sites do not relate to the methodologies employed by Dr. Sturchio to analyze Pomona's groundwater. EPA's warning regarding the application of isotope analysis to new, untested areas is a valid basis to require additional indicia of reliability for those new areas of application. See Att'y Gen. of Okla. v. Tyson [*1046]  Foods Inc., 565 F.3d 769, 780-81 (10th Cir. 2009)  [**19] (rejecting a new application of PCR method DNA typing where there was no testing or peer-reviewed publications specific to the application). In this case, however, the stable isotope study of chlorine and oxygen in perchlorate found in groundwater has been tested, analyzed, and subjected to peer review for at least ten years.

Thus, despite the fact that there is no EPA-certified method of analysis, the record shows that Dr. Sturchio's methodology and report are based on the scientific method, practiced by recognized scientists in the field, and have a basis in the knowledge and experience of the relevant discipline, thereby rendering the report reliable. See Southland Sod Farms, 108 F.3d at 1141. Dr. Sturchio's expert report details how he analyzed the relevant data and applied the data to reach his conclusions. The Federal Rules of Evidence do not require an endorsement from the EPA approving Dr. Sturchio's results. The district court's conclusion to the contrary was an abuse of discretion. See Preston, 706 F.3d at 1118.

3. Testing and retesting

The district court also excluded Dr. Sturchio's testimony because his methods "have not been tested by other laboratories and are not subject  [**20] to retesting given the failure to take dual samples." In order for a scientific technique to be reliable, there must be evidence in the record indicating the methodology "can be or has been tested." Cooper v. Brown, 510 F.3d 870, 880-81 (9th Cir. 2007). The question is whether an expert's methodology can be "challenged in some objective sense, or whether it is instead simply a subjective, conclusory approach that cannot reasonably be assessed for reliability." Fed. R. Evid. 702 Advisory Committee's Note to 2000 Amendments. Daubert described the "testability" prong in the context of a hypothesis that is falsifiable. 509 U.S. at 593. Testability "assures the opponent of proffered evidence the possibility of meaningful cross-examination (should he or someone else undertake the testing)." United States v. Mitchell, 365 F.3d 215, 238 (3d Cir. 2004). The district court incorrectly applied this standard.

The district court's conclusion was erroneous for three reasons: (1) other laboratories have tested the methodologies from the Guidance Manual used by Dr. Sturchio; (2) Dr. Sturchio's procedures are subject to retesting by another laboratory; and (3) challenges to the results obtained by using the techniques from the Guidance Manual go to the weight of the evidence and are a question for the fact finder, not the trial court.

First, Dr. Sturchio's methods were fully disclosed in the Guidance Manual and are the same methods that Dr. Sturchio used in his analysis of Pomona's groundwater.⁴ The Guidance Manual represents the latest compilation of QA/QC processes for any laboratory engaged in stable isotope analysis of perchlorate. The test under Daubert is whether the method "can be or has been tested." Cooper, 510 F.3d at 880 (citation omitted). In Cooper, for example, the court excluded expert testimony because there was "no evidence in the record that application of mass spectrometry to forensic analysis of blood evidence to determine EDTA levels can be or has been tested." Id. Unlike in Cooper, here several laboratories have used and tested the methodologies described in the Guidance  [*1047]  Manual, including the U.S. Geological Survey, the Oak Ridge National Laboratory, and the University of Illinois at Chicago where Dr. Sturchio works. Although Dr. Sturchio operates the only commercial laboratory using this methodology, testing at governmental laboratories demonstrates that Dr.  [**22] Sturchio's methods can be objectively challenged.

4   SQMNA contends that Dr. Sturchio's methods are not fully disclosed. Dr. Sturchio, however, provided a detailed description of the Pomona analysis in his expert report, which correlates with the processes described in the Guidance Manual.

Second, Dr. Sturchio's processes are subject to retesting. Under Daubert's testability factor, the primary requirement is that "[s]omeone else using the same data and methods . . . be able to replicate the result[s]." Zenith Elecs. Corp. v. WH-TV Broad. Corp., 395 F.3d 416, 419 (7th Cir. 2005). The district court stated that the "failure to take dual samples" meant that Dr. Sturchio's "methods" could not be retested. SQMNA argues that the district court did not err because there were two relevant defects in Dr. Sturchio's sampling procedures: (1) Dr. Sturchio failed to use duplicate columns in collecting groundwater samples; and (2) Dr. Sturchio failed to take split samples in order to compare analytical results.

SQMNA's defense of the district court's ruling is unpersuasive because both grounds for exclusion are without adequate support in the record. Neither of the alleged "defects" are "required" analytical  [**23] steps for stable isotope analysis and, hence, neither are necessary for retesting to occur. The use of duplicate columns during sampling is not mandatory. The basic diagram of the technique employed by Dr. Sturchio shows that the duplicate ion exchange column is "optional." The Guidance Manual also explains that "[i]n many instances, single IX columns are collected from each well." Duplicate columns are recommended for use on wells that have low levels of perchlorate, not for all sampling. In addition, the sample splitting mentioned in the Guidance Manual also is not mandatory. Dr. Sturchio contends that he ran duplicate analyses of his samples, verifying the Pomona results. Dr. Sturchio's Pomona results were also consistent with the pre-litigation Chino Basin Watermaster study. SQMNA correctly notes that Dr. Sturchio failed independently to verify his test results with a separate lab. This point, however, may serve to undermine or impeach the weight that should be afforded to Dr. Sturchio's testimony, but it does not refute the scientific reliability of his analysis.

Third, it is a question for the jury, not the court, to determine what weight to afford Dr. Sturchio's testimony. SQMNA argues that Pomona did not follow the Guidance Manual protocols and that Pomona's collection and extraction procedures were "makeshift." Daubert, however, "does not forbid admission" of a report where the weight of the conclusions are subject to challenge. United States v. Brannon, 146 F.3d 1194, 1196 (9th Cir. 1998) (permitting the admission of breathalyzer evidence where the scientific technique was not challenged, but rather, the results obtained).

SQMNA's argument relates to adherence to protocol, which typically is an issue for the jury. See Chischilly, 30 F.3d at 1154. SQMNA urges the Court to take a guarded approach to the issue of an expert's adherence to protocol. See, e.g., In re Paoli R.R. Yard PCB Litig., 35 F.3d 717, 745 (3d Cir. 1994) (holding that "any step that renders the analysis unreliable . . . renders the expert's testimony inadmissible. This is true whether the step completely changes a reliable methodology or merely misapplies that methodology."). In the Ninth Circuit, however, expert evidence is inadmissible where the analysis "is the result of a faulty methodology or theory as opposed to imperfect execution of laboratory techniques whose theoretical foundation is sufficiently accepted in the  [*1048]  scientific community to pass muster under Daubert." Chischilly, 30 F.3d at 1154 & n.11 (citations omitted). The rationale of this approach is that "[a] minor flaw in an expert's reasoning or a slight modification of an otherwise reliable method" does not render expert testimony inadmissible. Amorgianos v. Nat'l R.R. Passenger Corp., 303 F.3d 256, 267 (2d Cir. 2002). A more measured approach to an expert's adherence to methodological protocol is consistent with the spirit of Daubert and the Federal Rules of Evidence: there is a strong emphasis on the role of the fact finder in assessing and weighing the evidence. Daubert, 509 U.S. at 594-95.

The district court did not provide an explanation as to why Dr. Sturchio's alleged failure to adhere to the protocols in the Guidance Manual were significant enough to render his entire analysis unreliable. SQMNA argued to the district court that there was insufficient documentation of the sampling and extraction procedures. Dr. Sturchio's testimony, however, belies this conclusion. He explained that he had documentation verifying that the sampling procedures were followed pursuant to the Guidance Manual. He also verified  [**26] in his expert report and during the Daubert hearing that he followed the very detailed standard operating procedure for every sample that was analyzed. The district court did not apply the correct rule of law: only a faulty methodology or theory, as opposed to imperfect execution of laboratory techniques, is a valid basis to exclude expert testimony. Chischilly, 30 F.3d at 1154. Ignoring a controlling rule of law constitutes an abuse of discretion. See Preston, 706 F.3d at 1118. Moreover, given that Dr. Sturchio refuted SQMNA's assertion that the Guidance Manual protocols were not followed, the district court's application of the Chischilly standard is "without . . . inferences that may be drawn from the facts in the record." See id.

SQMNA's arguments challenging Dr. Sturchio's expert testimony are not uncontroverted, and they go to the weight that a fact finder should give to his expert report. The district court erroneously ruled that Dr. Sturchio's methodologies have not been and cannot be tested.

4. Reference database

The district court ruled that Dr. Sturchio's "reference database is too limited in order for him to reliably comment on the exclusiveness of the location of the potential  [**27] source of perchlorate in Pomona's water with an acceptable rate of error." The district court, however, was presented with conflicting expert evidence. SQMNA's expert Dr. Ramon Aravena ("Dr. Aravena") contended that the perchlorate reference database was too small. Dr. Sturchio, on the other hand, explained that the database was sufficiently large to permit him reasonably to draw a connection to the Atacama perchlorate.

At the Daubert hearing, the trial court was presented with Dr. Sturchio's analysis that the "dominant source of perchlorate in the Pomona groundwater is from Atacama (Chile)" and that the samples contained "minor amounts of perchlorate from other non-Atacama sources including synthetic and/or indigenous natural sources." Dr. Aravena's expert report cautioned that "not all the potential perchlorate sources have been characterized." Dr. Sturchio, however, responded to Dr. Aravena's contention by arguing that Dr. Aravena's opinion was based on disclosures and quotations from old and outdated publications. Dr. Sturchio also explained that when the Pomona study was conducted, synthetic and Atacama sources of perchlorate were well known and well characterized. At most, this  [**28] battle among experts merely  [*1049]  shows that Dr. Sturchio may not know the isotopic composition of every source of perchlorate in the world with a certainty. Under Daubert, however, such a "certainty" is not required, thus making this an invalid basis to exclude expert testimony. 509 U.S. at 590.

The Supreme Court has stated that "[t]rained experts commonly extrapolate from existing data." General Elec. Co. v. Joiner, 522 U.S. 136, 146, 118 S. Ct. 512, 139 L. Ed. 2d 508 (2007). It is where expert opinion is "connected to existing data only by the ipse dixit of the expert" that there may be "too great an analytical gap between the data and the opinion preferred" to support inclusion of the testimony. Id. Joiner requires an expert to justify a foundational assumption or refute contrary record evidence.

Chischilly is illuminating on this point. In that case, the defendant challenged the use of the FBI's ethnic-specific database for Native Americans (the "I-3 database"). 30 F.3d at 1155. The court considered whether the FBI's I-3 database was too small and may have contained too few Navajos to be reliable. Id. Both sides of the debate could find "support in the journals and research, and both sides [had] prominent spokespeople."  [**29] Id. Under Daubert's liberal standard, this sort of debate functioned more as an adverse admission and proved deadlock on both sides of an issue. Id. at 1155-56. The Chischilly court found that this evidence disproved a lack of "general acceptance" in the scientific community. Id.

The Chischilly analysis also demonstrates how trial courts ought to treat conflicting expert testimony. A factual dispute is best settled by a battle of the experts before the fact finder, not by judicial fiat. Where two credible experts disagree, it is the job of the fact finder, not the trial court, to determine which source is more credible and reliable. United States v. Sandoval-Mendoza, 472 F.3d 645, 654 (9th Cir. 2006).

The district court's resolution of this debate was an abuse of discretion and sufficient grounds for reversal. See Preston, 706 F.3d at 1118. Under Rule 702, it is reasonable for the jury to be presented with conflicting expert testimony. Sandoval-Mendoza, 472 F.3d at 654. Even if Dr. Sturchio's conclusions were "shaky," they should be attacked by "cross examination, contrary evidence, and attention to the burden of proof, not exclusion." Primiano, 598 F.3d at 564. The district court abused  [**30] its discretion in concluding that the reference database was too small. This is a matter for the jury.