A jury convicted the defendant, Carlos Fernandez, of possession of cocaine with intent to distribute, G. L. c. 94C, § 32A (c); and for doing so within one hundred feet of a public park, G. L. c. 94C, § 32J.1 At a subsequent jury-waived trial, the defendant’s convictions were found to be subsequent offenses. G. L. c. 94C, § 32A (d). We granted the defendant’s application for direct appellate review in order to consider whether it was error to admit certain evidence at trial, including evidence seized from the defendant’s automobile that was parked in a driveway adjacent to his residence, the results of field drug tests, and laboratory certificates of drug analysis. We conclude that it was not error to admit the first two categories of evidence, but that the erroneous admission of the certificates requires reversal of the defendant’s convictions and remand for a new trial.
1. Evidence at trial. In early 2005, members of a drug task force began surveilling the defendant. On seven occasions, undercover police officers collected bags of trash left on the street outside the defendant’s residence. The officers were able to identify the defendant’s trash based on the presence of letters and bulk *139mail addressed to him. Each week, small plastic bags were discovered in the defendant’s trash.2 The comers of many of the plastic bags had been cut off, consistent with drug distribution, and many of the plastic bags contained a white powder residue that was visually consistent with cocaine.3 On multiple occasions, undercover officers observed the defendant being driven around in various vehicles and engaging in behavior consistent with drug distribution, including performing hand-to-hand transactions with other individuals.
The jury heard testimony that the police conducted tests on the residue of at least one plastic bag pulled from the defendant’s trash each week. The procedure was described as a “presumptive test” or a “field test” designed to detect the presence of cocaine through the use of a chemically treated swab that turned from pink to blue in the case of a positive result.4 *All the field tests performed on the plastic bags found in the defendant’s trash returned positive results for the presence of cocaine.5 After testing, the police discarded the testing swabs.6
In addition to performing field tests, the police sent all the plastic bags in which white powder residue was visible ■— forty-seven in total — to be analyzed by a State criminal laboratory. The laboratory aggregated the residue from all forty-seven bags to obtain a single, testable sample. That is, it did not test each plastic bag individually for the presence of cocaine. Through the testimony of the police officer who gathered the plastic bags to be analyzed, the Commonwealth introduced a certificate of *140analysis from the Department of Public Health stating that the aggregated residue from the forty-seven plastic bags tested positive for cocaine.7
In March, 2005, the police obtained a warrant to search the defendant’s apartment. As the police arrived at the residence to perform the search, the defendant pulled into the driveway in a Toyota automobile registered to his girl friend. He was placed in handcuffs, and a cellular telephone, $1,493 in cash, and keys to both the Toyota and a Honda automobile were taken from his person. The Honda, known to be registered to the defendant, was parked in the driveway alongside the home.
The ensuing search of the apartment turned up little evidence of drug use or distribution. While large sums of money were discovered, as were unused latex gloves and plastic bags of the kind that had been found in the defendant’s trash, no cocaine or implements of the drug trade, such as scales or diluting materials, were found.
The police then turned their attention to the vehicles parked in the driveway of the residence, the Toyota and the Honda. A search of the Honda produced two pieces of inculpatory evidence: a small plastic bag containing one-half gram of cocaine and a black “magic marker” that had been hollowed out so that small items could be concealed within its cavity.8 A second *141certificate of analysis was introduced at trial confirming that the substance in the bag was cocaine. The defendant did not object to the introduction of the certificate.
In closing argument, defense counsel emphasized that the evidence that the defendant had distributed cocaine, rather than merely possessed it, was weak. For example, he argued that the aggregation technique used to test the defendant’s trash did not indicate that drugs were present each week. The prosecutor countered that the direct evidence that was recovered supported the inferences available from the field tests and the laboratory tests, namely that the defendant had possessed drugs with an intent to distribute them.
2. Claims of error. The defendant contends that the most damning evidence against him — the fruits of the search of the Honda, the field test testimony, and the certificates of analysis from the laboratory — should not have been admitted at trial. We take each up in turn.
a. Search of defendant’s Honda. The police obtained a warrant to search the defendant’s residence, the first-floor apartment of the building. At the time of the search, a blue Honda automobile, known to be registered to the defendant, was parked alongside the building in a narrow driveway, but the search warrant did not include the automobile expressly as a location to be searched.* ***9 At some point, the defendant’s girl friend, who was present during the search of the residence, informed the police that the defendant used the Honda to distribute narcotics and that there was a concealed compartment in the car used to hide drugs. Acting on this information, the police searched the *142automobile and recovered the hollowed out marker and the small bag of cocaine.
Prior to trial, the defendant filed a motion to suppress the evidence obtained from the search of the Honda, arguing that the search warrant for his apartment did not extend to this vehicle because it was not within the “curtilage” of the apartment.10 The motion judge denied the motion, concluding that the warrant’s scope included the vehicle because the driveway where it was parked was “intimately tied” to the defendant’s residence.11 The defendant filed a motion for reconsideration, which the judge allowed, but which was followed by a renewed denial of the defendant’s underlying motion to suppress.
“In reviewing a ruling on a motion to suppress, we accept the judge’s subsidiary findings of fact absent clear error ‘but conduct an independent review of [the judge’s] ultimate findings and conclusions of law.’ ” Commonwealth v. Scott, 440 Mass. 642, 646 (2004), quoting Commonwealth v. Jimenez, 438 Mass. 213, 218 (2002). In the context of a curtilage determination, we undertake our independent review cognizant that there is no “finely tuned formula” that demarcates the curtilage in a given case. See Commonwealth v. McCarthy, 428 Mass. 871, 874 (1999) (McCarthy), quoting United States v. Dunn, 480 U.S. 294, 301 (1987) (Dunn). A “ ‘correct’ answer to all extent-of-curtilage questions” is not always available. Dunn, supra. This is especially true when multi-unit dwellings with various “common areas” are at issue. See Commonwealth v. Dora, 57 Mass. App. Ct. 141, 144 (2003) (“Whether [an] occupant of a multi-unit apartment building has a reasonable expectation of privacy [in common areas] is a question that cannot be answered categorically”); United States v. Stanley, 597 F.2d 866, 870 (4th *143Cir. 1979) (“ ‘common area’ curtilage issue has been a thorny one for the courts”).
The defendant argues that two of our cases, Commonwealth v. Thomas, 358 Mass. 771, 774-775 (1971); and McCarthy, supra, stand for the proposition that the driveway of a multiunit dwelling cannot come within the curtilage of any individual unit. See Commonwealth v. Thomas, supra (“In a modem urban multi-family apartment house, the area within the ‘curtilage’ is necessarily much more limited than in the case of a rural dwelling subject to one owner’s control”). But the Thomas and McCarthy decisions did not establish a bright line mle that shared or common driveways connected to multi-unit dwellings are never within the curtilage of one of the units. There is no such bright line. See United States v. Diehl, 276 F.3d 32, 39 (1st Cir.), cert, denied, 537 U.S. 834 (2002). Rather, it is the policy of the United States Supreme Court and of this court to approach curtilage questions on a case-by-case basis. See Dunn, supra; McCarthy, supra. If a vehicle parked in a driveway is within the curtilage, it may be searched pursuant to a warrant issued to search the accompanying residence. Commonwealth v. Signorine, 404 Mass. 400, 403 (1989).
In determining whether a particular parking area is within the curtilage of a residence, we consider four factors: “(1) the proximity of the area to the home, (2) whether the area is included within an enclosure surrounding the home, (3) the nature of the uses to which the area is put, and (4) the steps taken by the resident to protect the area from observations by people passing by.” McCarthy, supra, citing Dunn, supra. However, these four factors (Dunn factors) cannot be “mechanically applied.” Dunn, supra. They are “useful analytical tools only to the degree that, in any given case, they bear upon the centrally relevant consideration — whether the area in question is so intimately tied to the home itself that it should be placed under the home’s ‘umbrella’ of . . . protection [under the Fourth Amendment to the United States Constitution].” Id.
McCarthy involved the search of an automobile parked in a parking space of a common parking lot that could accommodate fifty vehicles. McCarthy, supra at 872 n.4. The police had obtained a search warrant only for the defendant’s apartment. *144Id. at 872. The parking space in McCarthy was twenty-five feet from the entrance of the apartment building; it was a space assigned to visitors of the building; and the parking lot itself frequently was used by tenants, visitors, delivery people, maintenance workers, “and anyone else with business at the building.” Id. at 872, 875. The parking lot was large, and it was free from any landscaping or other design features that made any single parking space more private than the others. Id. at 875 & n.5.
Applying the four Dunn factors, we concluded that the space in which the defendant parked his automobile was not proximate to his apartment; the space was not enclosed in any manner; the space was used by many individuals without exclusion; and the defendant took no steps to protect the space from observation. Id. at 875-876. In particular, we noted that the defendant had parked in a space designated for visitors, rather than his assigned space. Id. at 872 & n.4. For the parking space to be within the curtilage of the defendant’s apartment in these circumstances would have required us to adopt a “roaming zone of curtilage that would follow the defendant’s car to whichever parking space he select[ed] at the apartment complex on a given day.” Id. at 874. That we would not do, concluding that “neither the physical layout of the parking lot, nor the nature of its regular and intended uses, would give tenants any reasonable expectation that it would be ‘treated as the home itself.’ ” Id. at 875, quoting Dunn, supra at 300.
The present case differs in significant ways. Here, the defendant lived in the first-floor apartment of a “three-family home,” not an “apartment complex.” Directly adjacent to the house was a narrow driveway, approximately the width of one vehicle and the length of two, as compared to a lot that could accommodate fifty automobiles. There is no evidence that there was a side door or other way to gain access to the driveway from any of the other apartments. No one seeking to enter or visit the home would have had reason to traverse the driveway. Additionally, during the course of their surveillance, the police officers saw only vehicles associated with the defendant — his blue Honda, for one — parked in the driveway.
Applying the Dunn factor analysis to the facts found in this *145case supports the motion judge’s conclusion that the driveway in the present case was within the curtilage of the defendant’s apartment.
i. Proximity. The driveway was highly proximate to the defendant’s apartment, both in absolute terms (it bordered it) and in relative terms (the driveway was closer to the defendant’s first-floor apartment than to the other apartments). In McCarthy, we emphasized that between the parking space and the defendant’s apartment “presumably lay common hallways, entrances, and exits,” indicating that the parking space was attenuated from the apartment. McCarthy, supra at 875. The record in the present case indicates that the house had a common front door, through which individuals entering and leaving the apartments would pass. But the attenuation caused by such common passage is significantly less than was present in McCarthy, and the motion judge found that tenants and visitors of the defendant’s building would have had no reason to traverse the driveway to enter. Passage through common areas was required to reach the driveway, but passage through the driveway was not required to reach the common areas.
ii. Enclosure. In the context of curtilage determinations, whether an area is enclosed is a question whether it is marked in such a fashion that connects it to or encompasses it within the curtilage of the home in question. See Dunn, supra at 302 (area searched was not within fence surrounding main house); McCarthy, supra at 875 (parking lot enclosed, but enclosed to all tenants generally). Here, the photographic evidence submitted during the hearing on the motion to suppress shows a fence along the driveway separating it from the neighboring building and connecting it to the defendant’s building.12 It was thus enclosed in relation to the building itself, but nothing indicated that the driveway was enclosed to the defendant’s individual apartment. In the language of the United States Court of Appeals for the Fourth Circuit, the enclosure of the driveway in this case did not show that the driveway was “annexed to” the defendant’s *146apartment. United States v. Stanley, 597 F.2d 866, 870 (4th Cir. 1979). However, compared to the circumstances in McCarthy, the driveway was enclosed to far fewer possible tenants, and no visitor parking existed.
iii. Use. In Dunn, the Court discussed the fact that the area in question — a bam some distance away from the main house — was not being used as an extension of the home. Id. at 302-303. The police smelled strong chemical odors and heard engines running inside the bam, suggesting that the bam was not being put to domestic use and cutting against its inclusion within the curtilage of the dwelling. Id. In McCarthy, supra at 875, we emphasized that the parking lot was used by “anyone . . . with business at the building.” As in Commonwealth v. Thomas, 358 Mass. 771, 774 (1971), the defendant in McCarthy exercised no control over the parking lot, or even the space in which he parked.
Here, the motion judge found that the driveway was not used by nontenants because it was not necessary to traverse the driveway to enter or exit the building. He also credited testimony that the police officers did not observe anyone using the driveway except the defendant and individuals associated with him. In sum, all the evidence at the hearing pointed to the driveway belonging to the defendant’s exclusive use.
iv. Steps taken to protect from observation. There is no evidence in this case that the defendant took any steps to conceal the Honda from observation. The automobile was visible from the street. See Commonwealth v. Simmons, 392 Mass. 45, 46-49, cert, denied, 469 U.S. 861 (1984). However, in order to see into the car, an observer would have been required to depart from the path used to reach the front door, and travel along the side of the house down the driveway.
Accordingly, pursuant to the “well-settled principle . . . that the scope of a warrant authorizing the search of a particularly described premises, includes automobiles owned or controlled by the owner thereof, which are found on the premises,” Commonwealth v. Signorine, 404 Mass. 400, 403 (1989), the police were authorized to search the defendant’s Honda.13 The evidence that resulted was properly admitted.
*147b. Field tests. The defendant also contends that it was error to permit the Commonwealth to elicit testimony about the field tests performed by the investigating police officers on individual plastic bags found in the defendant’s trash over several weeks. He argues that the Commonwealth was permitted to make use of the field tests without first establishing that they were reliable. Commonwealth v. Lanigan, 419 Mass. 15, 26 (1994) (Lanigan).14
The defendant first argued for the necessity of a Lanigan hearing on the initial day of trial, several months after the Commonwealth alerted him about the field testing and how it was *148performed. The judge declined to hold a full evidentiary hearing to determine if the field tests, as offered, were reliable. However, he did require the Commonwealth to provide “some foundation” establishing the field tests’ reliability before the presentation of the field test evidence to the jury. The Commonwealth responded by supplying the judge with a case from the Criminal Court of the City of New York, People v. McIntyre, 185 Misc. 2d 58 (N.Y. Crim. Ct. 2000), and a regulation applicable to the Department of Correctional Services of the State of New York, N.Y. Comp. Codes R. & Regs. tit. 7, § 1010.5 (2006), both of which discuss the general acceptance of field tests (of the same brand used in the defendant’s case) for obtaining presumptive results.15 Based on these submissions, the judge ruled that the field tests were sufficiently reliable and that the Commonwealth could elicit testimony about their results.
On this record, the defendant’s first argument — that the judge did not subject the field tests to a Lanigan analysis —■ is without basis.16 The judge did perform such an analysis, albeit a limited one as a consequence of the defendant’s decision to wait until the day of trial to raise the issue. The more relevant question in this case is whether the Lanigan analysis that was performed was adequate.
“[A] party seeking to introduce scientific evidence may lay *149an adequate foundation ... by establishing general acceptance in the scientific community.” Canavan’s Case, 432 Mass. 304, 310 (2000). Our review is for abuse of discretion in order to “allow trial judges the needed discretion to conduct the inherently fact-intensive and flexible Lanigan analysis, while preserving a sufficient degree of appellate review to assure that Lanigan determinations are consistent with the law and supported by a sufficient factual basis in the particular case.” Id. at 312.
The defendant’s position that the Lanigan analysis was inadequate rests on a mischaracterization of how the Commonwealth intended to use, and did eventually use, the field test evidence at trial. Field tests, he asserts, are generally accepted only for “presumptive,” as opposed to conclusive, results. Only formal laboratory testing (insulated from the conditions of the field) can confirm the identity of a substance.17 Thus, he argues, the use of field tests as conclusive evidence of drug identity exceeds the purpose for which such tests are generally accepted and, therefore, a judge must subject field tests to a Lanigan analysis before permitting the Commonwealth to deploy them as anything more than presumptive results.
The defendant argues that the judge ran afoul of Lanigan by permitting the Commonwealth to elicit testimony about individual field tests that were never individually confirmed by a laboratory. Rather, he reminds us, the laboratory aggregated the residue collected from all of the defendant’s trash before performing its analysis. As a result, the field tests were the only evidence that established that cocaine was present in the defendant’s trash on multiple weeks (indicative of drug distribution), which, in the defendant’s view, transformed the field tests into conclusive evidence. We disagree with this characterization.
Here, two factors weigh against concluding that the judge abused his discretion in this case. First, as noted, we cannot overlook that the defendant waited until the eleventh hour to raise the Lanigan issue. Because of the short notice, with trial *150scheduled to start that day, the Commonwealth could not call a scientist from the field test manufacturer to testify as to its reliability. Second, despite the defendant’s depiction of how the field tests were used at trial, neither the Commonwealth nor the judge ever characterized the tests as conclusive. To the contrary, the “presumptive” nature of the tests was emphasized at every turn.18 The judge twice instructed the jury that:
“A single field test swab of an object may or may not yield a valid result. No chemical reagent system adaptable for field use exists that will completely eliminate the occurrence of an invalid test result. A forensic laboratory test is necessary to identify the nature of an unknown substance completely.”19
In this light, it was not an abuse of discretion for the judge, acting on short notice, to accept the Commonwealth’s relatively weak proffer (the New York authorities) as establishing that field tests were sufficiently reliable for a limited purpose. The reliability of scientific evidence is distinct from its strength. See Commonwealth v. McNickles, 434 Mass. 839, 854 (2001) (“That a scientific test provides descriptive information short of a definitive identification does not make it unreliable, irrelevant, or otherwise inadmissible”); Commonwealth v. Duguay, 430 Mass. 397, 402 (1999) (jury warned of possible false positive results). Moreover, judges may rely on the decisions of other *151courts as one factor when evaluating the general acceptance of scientific evidence. See Commonwealth v. Mendes, 406 Mass. 201, 209 (1989); Commonwealth v. Whynaught, 377 Mass. 14, 17-18 (1979).20 The defendant was entitled to use the weakness of the field test evidence against the Commonwealth, and he did so aggressively.
c. Motion to strike field test testimony. After cross-examining the detective about the field tests, the defendant moved to strike the testimony on the ground that the tests were not performed properly. The motion was denied. The defendant argues that the denial of his motion to strike erroneously absolved the Commonwealth of its burden “to show that the testifying expert properly performed a scientifically valid methodology in arriving at his opinion.” Commonwealth v. Vao Sok, 425 Mass. 787, 798 (1997). We disagree.
Prior to the detective’s testimony, there was very little suggestion that his testing techniques were suspect.21 On cross-examination, however, the detective admitted that he swabbed each plastic bag only once. According to the New York State regulation that the judge relied on to establish the field tests’ reliability as a presumptive test, a single swab was less reliable *152than multiple swabs.22 On that basis, the defendant argued that the detective’s testimony revealed that his testing was not reliably performed and therefore that the field test evidence should be struck from the record.
The judge did not abuse his discretion in denying the defendant’s motion to strike. The regulatory language on which the defendant bases his argument does not establish that the officer performed invalid tests. Rather, the language establishes that single swab field tests are even more inconclusive than multiple swab field tests. This was consistent with the manner in which the evidence was presented to the jury, and was emphasized by the judge in limiting instructions.
d. Certificates of drug analysis. Finally, the defendant argues that he must be afforded a new trial because the certificates of analysis establishing that cocaine was present in his trash and in the plastic bag recovered from his automobile were introduced in violation of this right to confront witnesses against him.
Prior to his trial, this court had held that certificates of analysis were not “testimonial” statements within the meaning of the confrontation clause of the Sixth Amendment to the United States Constitution. Commonwealth v. Verde, 444 Mass. 279, 284 (2005). Accordingly, at the time of his trial, the Commonwealth was permitted to admit the certificates of analysis without calling the technicians who performed the laboratory tests to testify. After the defendant was convicted, the United States Supreme Court held that certificates of analysis are testimonial statements whose admission in evidence must be accompanied by live testimony that can be confronted. Melendez-Diaz v. Massachusetts, 129 S. Ct. 2527, 2532 (2009). Since then, in Commonwealth v. Vasquez, 456 Mass. 350, 352, 356-358, 360 (2010), we have held that individuals such as the defendant — whose trials took place after our decision in Verde, but prior to *153the granting of certiorari by the Supreme Court in Melendez-Diaz — should not be penalized for their failure to object to the introduction of certificates of analysis. Instead, even where certificates were admitted without objection, we consider “whether admission of the drug certificates at trial was harmless beyond a reasonable doubt.” Commonwealth v. Vasquez, supra at 360.23
The present case must be remanded for a new trial because we cannot say that “the guilty verdict[s] actually rendered in this trial [were] surely unattributable to the error” (emphasis in original). Commonwealth v. Vasquez, supra at 361, quoting Sullivan v. Louisiana, 508 U.S. 275, 279 (1993). We begin with the certificate of analysis attesting to the fact that the aggregated trash residue contained cocaine. Although we emphasized the absence of field tests in Commonwealth v. Vasquez, supra at 364, their presence in this case does not alter the result. Our discussion concerning the limited purpose for which the field test evidence was offered demonstrates that the tainted evidence — the certificates — would have had an effect on the jury. Id. at 360, quoting Commonwealth v. Tyree, 455 Mass. 676, 701 (2010). Indeed, the judge’s description of the field tests as only “incremental” evidence, and his pointing to the necessity of a forensic laboratory test to identify the substance “completely,” solidify our conclusion that the certificate played an important role in proving the identity of the substance contained in the defendant’s trash.
As to the certificate of analysis admitted to prove that the plastic bag found under the seat of the defendant’s automobile was cocaine, it was the only evidence of the identity of the bag’s contents. Without question, the admission of the certificate of analysis was not harmless beyond a reasonable doubt.
*1543. Conclusion. Because the admission of the two certificates of analysis was not harmless beyond a reasonable doubt, we reverse the judgments, set aside the verdicts, and remand the cases for a new trial. On retrial, if the Commonwealth seeks to introduce field test evidence, the judge should engage in a new Lanigan analysis.
So ordered.
The defendant was tried with Thomas Catapano, who was convicted of possession of marijuana but is not a party to this appeal.
One of the trash searches alone resulted in the discovery of more than one hundred plastic bags stuffed inside a single latex glove.
Empty boxes of plastic bags, numerous air fresheners, and notations containing calculations and dollar amounts were also discovered.
The tests, referred to as the “Scott-Reagent tests,” were manufactured by NIK Public Safety, Inc., and an unused box of testing swabs was introduced in evidence.
At the defendant’s request, the trial judge told the jury, both during trial and after the close of the evidence, that a single field test swab of an object might not yield a valid result, that invalid results were possible, and that forensic laboratory tests were necessary to confirm with accuracy the identity of a substance.
Defense counsel impeached the procedures followed by the police officers. For example, they did not file an incident report after any of the field tests, and “lumped” all the discarded plastic bags gathered on different days together into a single container.
The defendant objected to the introduction of the certificate, arguing that the certificate’s language was phrased in a way that suggested that each of the plastic bags contained cocaine, when, in fact, the laboratory had aggregated the contents of all forty-seven bags before performing any tests. The defendant asked that the certificate be redacted in a way inconsistent with that suggestion or, alternatively, that the Commonwealth stipulate that the laboratory had aggregated the samples before testing. The judge permitted the unredacted certificate to be introduced in evidence, provided that the Commonwealth elicit testimony from the detective about the aggregation technique that was used by the laboratory.
The defendant did not object that the detective was not competent (for lack of personal knowledge) to testify about how the laboratory conducted its tests. Rather, he argued that the lack of competence could be remedied by the Commonwealth stipulating that the aggregation technique was used. The defendant inveighed against the aggregation technique in his closing argument.
The defendant did not object to the introduction of the certificate on the basis of his right to confront witnesses against him. See Commonwealth v. Vasquez, 456 Mass. 350, 360 (2010).
The cocaine was wrapped in the comer of a plastic bag. A detective testi*141tied that packaging cocaine in such a manner was consistent with drug distribution, but he also admitted that the quantity recovered in the Honda was consistent with personal use. The detective also testified that hollowed out markers were “almost perfect to carry a certain amount of drugs for sale.”
The Honda automobile was mentioned in the affidavit submitted in support of the application for the warrant. The affidavit stated that this vehicle belonged to the defendant, and included it in the list of vehicles used by the defendant during hand-to-hand transactions. This likely was sufficient to establish probable cause to search the Honda, and had the police requested the automobile’s inclusion in the search warrant, their request likely would have been granted. “We note that this problem could have been avoided altogether had the police included the vehicle in the application for the search warrant.” Commonwealth v. McCarthy, 428 Mass. 871, 872 n.3 (1999).
“The curtilage concept originated at common law to extend to the area immediately surrounding a dwelling house the same protection under the law of burglary as was afforded the house itself.” United States v. Dunn, 480 U.S. 294, 300 (1987). It has since been applied to define those areas associated with a dwelling house that the police may not search without a warrant. Commonwealth v. McCarthy, supra at 873-874. In a “unique twist in the law,” United States v. Stanley, 597 F.2d 866, 870 (4th Cir. 1979), “[i]n the present context . . . curtilage serves a different function —• it helps define where the police can search pursuant to a warrant.” Commonwealth v. McCarthy, supra at 874.
The motion judge was not the trial judge.
Based on the photographic exhibits that the motion judge appended to his decision, it is clear that the driveway was a narrow alley between two buildings. Its width was defined by the exterior walls of the adjacent buildings, creating a sense that it was walled in on two sides.
As an alternative ground for denying the defendant’s motion to suppress, *147the motion judge suggested that the search of the automobile was lawful because the police had probable cause to search it after the defendant’s girl friend told them that he used the automobile to transport narcotics. We have not previously addressed whether the automobile exception to the warrant requirement may justify a search of an automobile parked within the curtilage of a defendant’s home. Given our conclusion that the Honda in this case came within the scope of the search warrant, it is unnecessary to decide this issue today. However, we leave open the possibility that, in certain circumstances, police officers may search a vehicle parked in a private driveway on the basis of probable cause alone. Cf. Commonwealth v. A Juvenile (No. 2), 411 Mass. 157, 160-162 (1991) (no expectation of privacy from police observation of automobile parked in private driveway where police entered driveway as any ordinary visitor would; seizure of automobile permitted on probable cause); Commonwealth v. Simmons, 392 Mass. 45, 46-49, cert, denied, 469 U.S. 861 (1984) (law enforcement permitted to enter onto private driveway used by all visitors to home to observe outwardly visible areas of automobile parked in driveway). We note that the exigencies associated with an automobile do not dissipate completely once the automobile is parked in a private driveway. See Commonwealth v. Nicholson, 58 Mass. App. Ct. 601, 607 n.7 (2003), quoting United States v. McCoy, 977 F.2d 706, 710 (1st Cir. 1992) (“inherent mobility of motor vehicles . . . and the reduced expectation of privacy associated with them . . . justify application of the vehicular exception ‘[ejven in cases where an automobile [is] not immediately mobile’ ” [citations omitted]); United States v. Brookins, 345 F.3d 231, 234, 237-238 (4th Cir. 2003), quoting Maryland v. Dyson, 527 U.S. 465, 466 (1999) (applying automobile exception to vehicle parked in private driveway because automobile was “readily mobile”).
In Commonwealth v. Lanigan, 419 Mass. 15, 25-26 (1994) (Lanigan), we reaffirmed that the admissibility of scientific evidence turns on the reliability of that evidence. Thus, prior to its admission, a judge “must rule first on any challenge to the validity of any process or theory underlying a proffered [scientific] opinion. ‘This entails a preliminary assessment of whether the reasoning or methodology underlying the testimony is scientifically valid and of whether that reasoning or methodology properly can be applied to the facts in issue.’ ” Id. at 26, quoting Daubert v. Merrell Dow Pharms., Inc., 509 U.S. 579, 592-593 (1993).
The case, People v. McIntyre, 185 Misc. 2d 58, 64 (N.Y. Crim. Ct. 2000), surveys the high degree of accuracy (in New York) of field tests made by the same manufacturer as the one used in the defendant’s case, and concludes that such field tests “provide a ‘reliable basis’ for inferring the presence of cocaine.”
The regulation, N.Y. Comp. Codes R. & Regs. tit. 7, § 1010.5(d) (2006), provides “a statement of the scientific principles and validity of the testing materials and procedures used” to field test suspected narcotics. The statement discusses the brand of field test used in the defendant’s case, and states that a “positive identification is considered a component of probable cause and generally recognized within our legal system as being presumptive in nature.” The statement is designed to be provided to an incarcerated inmate who is subjected to discipline after being found with narcotics.
The Commonwealth also supplied a second case, Davis v. State, 260 Ga. App. 853 (2003). Although referenced by the judge in his written ruling, all parties — counsel and the judge — agreed that it bore only indirectly on the issue.
In arguing that the judge should have subjected the field tests to a Lanigan analysis, the defendant relies on Commonwealth v. Sands, 424 Mass. 184, 188 (1997), and State v. Morales, 132 N.M. 146,149 (Ct. App. 2002). However, because the judge did perform a Lanigan review, these cases are inapposite.
Field tests for “presumptive identification” typically are followed by laboratory tests. The “near universal” laboratory test is the “gas chromatography-mass spectrometry” test. See National Research Council, Strengthening Forensic Science in the United States, A Path Forward 134-135 (2009). There is no information in the record of this case regarding the test or tests used by the State crime laboratory.
For example, in response to a question from the judge about how the field tests would be used at trial, the prosecutor stated:
“I don’t think that anyone is claiming that [the field test] is as perfectly reliable as the lab technique that’s used. For obvious reasons, someone who’s conducting surveillance on the street doesn’t have the ability to pull up a team of. . . technicians . . . but I think that it does serve the purpose for which the police were using it.”
For his part, the judge stated that the field tests were “cumulative or incremental evidence of the character of the controlled substance,” rather than conclusive evidence. The field tests, he noted, would be followed up with a certificate of analysis confirming that the aggregated sample from the trash contained cocaine, which in turn provided “some corroboration” that the field tests were reliable.
The instruction was derived from the New York State regulation used to establish the reliability of the field tests. N.Y. Comp. Codes R. & Regs. tit. 7, § 1010.5(d) (2006).
We caution that our conclusion that the judge did not abuse his discretion in this case is of limited scope. For one, it is based in part on the short notice given to the Commonwealth and the court in this case. Given more ample warning, a more searching Lanigan analysis would have been appropriate. Additionally, to date, no appellate case from Massachusetts has accepted as reliable field test results, regardless of the purpose for which they are offered. Until that occurs, field tests offered to prove the identity of a substance, “presumptive” or otherwise, must be evaluated according to one of the methods approved in Lanigan, supra at 26, such as a review for general acceptance. See Commonwealth v. Shanley, 455 Mass. 752, 763 n.15 (2010), citing Commonwealth v. Frangipane, 433 Mass. 527, 538 (2001). And, in general, determinations of general acceptance should be based on more than was proffered here. Moreover, as demonstrated by our discussion, infra, concerning the introduction of laboratory certificates of analysis after the decision of the United States Supreme Court in Melendez-Diaz v. Massachusetts, 129 S. Ct. 2527 (2009), the importance of field tests is likely to increase. See Commonwealth v. Vasquez, 456 Mass. 350, 364 (2010) (emphasizing absence of field test evidence).
Before the officer testified, the defendant raised the fact that one of the field tests may have been contaminated when the officer shook some residue onto a table before testing it. No claim was made about any of the other tests.
The regulation provides in relevant part:
“The results of a single test may or may not yield a valid result. However, the sequential results of several tests, if they all indicate a positive reaction for a particular substance, provides [sic] a high degree of certainty that the suspect material is in fact what the [field test] indicates it to be.”
N.Y. Comp. Codes R. & Regs. tit. 7, § 1010.5(d).
The defendant waited until March, 2010, just prior to oral argument before this court, to raise this issue, well after the decision in Melendez-Diaz v. Massachusetts, supra, was issued in June, 2009. Although we granted the defendant’s request to file an eleventh hour supplemental brief on the matter, given the fact that the defendant had the time and opportunity to include his Melendez-Diaz argument in his primary brief — indeed, he could have raised it in his application for direct appellate review, which he filed in July, 2009 — it is troubling that he waited. His timing prevented the Commonwealth from filing a response. We underscore that supplemental briefing is not a matter of right, and nothing in our decision in Commonwealth v. Vasquez, supra, sanctions the omission of an appellate issue that is no longer futile.