*1297This matter is before the court following the U.S. Department of Commerce's ("Commerce" or the "agency") redetermination upon remand in this case. See Final Results of Redetermination Pursuant to Court Remand ("Remand Results"), ECF No. 69-1.
Plaintiff Shenzhen Xinboda Industrial Co., Ltd. ("Plaintiff" or "Xinboda") initiated this action1 challenging Commerce's final results in the 20th administrative review ("AR 20") of the antidumping duty order on fresh garlic from the People's Republic of China ("PRC" or "China").2 See Summons, ECF No. 1; Fresh Garlic From the People's Republic of China , 81 Fed. Reg. 39,897 (Dep't Commerce June 20, 2016) (final results and final rescission of the 20th antidumping duty admin. review; 2013-2014) (" Final Results "), ECF No. 30-4, and accompanying Issues and Decision Mem., A-570-831 (June 10, 2016) ("I & D Mem."), ECF No. 30-5.3 Specifically, Xinboda, a mandatory respondent in this review, challenged Commerce's (1) rejection of surrogate country information demonstrating Mexico's economic comparability to China; (2) selection of Romania as the primary surrogate country; and (3) calculation of movement expenses. See Pl. Shenzhen Xinboda Industrial Co., Ltd.'s Mot. for J. on the Agency R., ECF No. 39, and Pl. Shenzhen Xinboda Industrial Co., Ltd. Mem. in Supp. of Mot. for J. on the Agency R. ("Xinboda's 56.2 Br."), ECF No. 39-2; I & D Mem. at 1. On December 18, 2017, the court remanded Commerce's rejection of surrogate country information and deferred consideration of Plaintiff's additional challenges pending the results of Commerce's remand redetermination. See *1298Shenzhen Xinboda Indus. Co. Ltd. v. United States , Slip Op. 17-160, 2017 WL 6502727 (CIT Dec. 5, 2017).4
On March 9, 2018, Commerce filed its Remand Results.5 On remand, Commerce, under protest,6 permitted Xinboda to submit factual information regarding Mexico's economic comparability to China. Remand Results at 1. Upon consideration of this information and Mexican surrogate value data, Commerce affirmed its selection of Romania as the primary surrogate country. Id. at 1, 31.
Xinboda filed comments opposing the Remand Results. See Pl. Shenzhen Xinboda Industrial Co., Ltd. Comments in Opp'n to U.S. Dep't of Commerce's Remand Redetermination ("Xinboda's Remand Opp'n"), ECF No. 84.7 Defendant United States ("Defendant" or the "Government") and Defendant-Intervenors8 filed comments in support of the Remand Results. See Def.'s Resp. to Comments on Remand Results ("Def.'s Remand Reply"), ECF No. 78; Def.-Ints.' Comments in Supp. of the U.S. Dep't of Commerce's Redetermination Pursuant to Remand ("Def.-Ints.' Remand Reply"), ECF No. 77.
For the reasons discussed herein, Commerce's selection of Romania as the primary surrogate country and selection of Romanian pricing data as the surrogate value for raw garlic are sustained. However, the court remands for further consideration Commerce's addition of transportation costs to the surrogate value for raw garlic and calculation of Xinboda's brokerage and handling and inland freight expenses.
JURISDICTION AND STANDARD OF REVIEW
The court has jurisdiction pursuant to § 516A(a)(2)(B)(iii) of the Tariff Act of 1930, as amended, 19 U.S.C. § 1516a(a)(2)(B)(iii) (2012),9 and 28 U.S.C. § 1581(c). The court will uphold an agency determination that is supported by substantial evidence and otherwise in accordance *1299with law. 19 U.S.C. § 1516a(b)(1)(B)(i). "The results of a redetermination pursuant to court remand are also reviewed for compliance with the court's remand order." SolarWorld Ams., Inc. v. United States , 41 CIT ----, ----, 273 F.Supp.3d 1314, 1317 (2017) (quoting Xinjiamei Furniture (Zhangzhou) Co. v. United States , 38 CIT ----, ----, 968 F.Supp.2d 1255, 1259 (2014) (internal quotation marks omitted).
DISCUSSION
I. Surrogate Country Selection
An antidumping duty is "the amount by which the normal value exceeds the export price (or the constructed export price) for the merchandise." 19 U.S.C. § 1673. When an antidumping duty proceeding involves a nonmarket economy country, Commerce determines normal value by valuing the factors of production10 in a surrogate country, see id. § 1677b(c)(1), and those values are referred to as "surrogate values." Commerce may also rely on surrogate values, when appropriate, to adjust the export price or constructed export price to account for costs incurred in "bringing the subject merchandise from the original place of shipment in the exporting country to the place of delivery in the United States." Id. § 1677a(c)(2)(A); see also Fine Furniture (Shanghai) Ltd. v. United States , 40 CIT ----, ----, 182 F.Supp.3d 1350, 1368 (2016) (noting Commerce's use of a surrogate value to calculate international movement expenses). In selecting surrogate values, Commerce must use "the best available information" that is, "to the extent possible," from a market economy country or countries that are economically comparable to the nonmarket economy country and "significant producers of comparable merchandise." 19 U.S.C. § 1677b(c)(1), (4).
Commerce generally values all factors of production in a single surrogate country.11 Commerce has adopted a four-step approach to selecting a primary surrogate country. See Import Admin., U.S. Dep't of Commerce, Non-Market Economy Surrogate Country Selection Process , Policy Bulletin 04.1 (2004), http://enforcement.trade.gov/policy/bull04-1.html (last visited Dec. 18, 2018) (hereinafter "Policy Bulletin 04.1"). Pursuant to Policy Bulletin 04.1,
(1) the Office of Policy ("OP") assembles a list of potential surrogate countries that are at a comparable level of economic development to the [non-market economy] country; (2) Commerce identifies countries from the list with producers of comparable merchandise; (3) Commerce determines whether any of the countries which produce comparable merchandise are significant producers of that comparable merchandise; and (4) if more than one country satisfies steps (1)-(3), Commerce will select the country with the best factors data.
Jiaxing Brother Fastener Co., Ltd. v. United States , 822 F.3d 1289, 1293 (Fed. Cir. 2016) (citation omitted); see also Policy Bulletin 04.1.
On remand, Commerce reopened the administrative record, invited Xinboda *1300to resubmit information regarding Mexico's economic comparability to China, and set deadlines for parties to rebut, clarify, or correct the information contained therein. Remand Results at 4; see also Refiling Resubmission of Sept. 17, 2015 Additional Surrogate Country List and Surrogate Country Comments (Jan. 3, 2018) ("Xinboda's SC Comments"), RR 6, PRJA Tab 10, Suppl. CRJA Tab 2, Suppl. PRJA Tab 2; Rebuttal to Xinboda'[s] Resubmitted Sept. 17, 2015 Additional Surrogate Country List and Surrogate Country Comments (Jan. 10, 2018) ("Xinboda's Remand Rebuttal"), RR 8-10, PRJA Tab 12; Pet'rs' Submission of Rebuttal Factual Information (Jan. 10, 2018) ("Pet'rs' Remand Rebuttal"), RR 11-16, PRJA Tab 13, Suppl. CRJA Tab 1, Suppl. PRJA Tab 1. Upon review of this information, Commerce determined that Mexico and Romania are both at the same level of economic development as the PRC and significant producers of comparable merchandise. See Remand Results at 8-9, 11.12 Pursuant to Policy Bulletin 04.1, Commerce's inquiry thus turned to which country provided the best factors data. Id. at 11.
To determine the best source of factors data, Commerce considers the degree to which the data is publicly available, contemporaneous with the period of review, tax and duty exclusive, representative of a broad-market average, and product-specific. Id. at 11-12; see also 19 C.F.R. § 351.408(c)(1), (4). Commerce's analysis of this issue is confined to the record built by interested parties. See QVD Food Co., Ltd. v. United States , 658 F.3d 1318, 1324 (Fed. Cir. 2011). Moreover, when selecting from among multiple sources, Commerce's inquiry may be a relative exercise and will rarely involve perfect data sets. When "Commerce is faced with the choice of selecting from among imperfect alternatives, it has the discretion to select the best available information for a surrogate value so long as its decision is reasonable." Catfish Farmers of Am. v. United States , 33 CIT 1258, 1273, 641 F.Supp.2d 1362, 1377 (2009). Applying the afore-mentioned criteria, Commerce weighed the information on the record and determined that the Romanian data was the best available information.
A careful review of the Remand Results indicates that Commerce's selection of Romania over Mexico was a close call. Publicly available data was available from both countries. See Remand Results at 12, 15, 30. Although roughly contemporaneous data was available from Mexico, Commerce considered the monthly data from Romania to be more contemporaneous than the annual data from Mexico, which would be impacted by months outside the period of review. Id. at 17. Data from both countries was also considered tax and duty exclusive. See id. at 12-13, 15, 27. Commerce expressly found that Romanian data represented a broad market average but made no such finding with respect to Mexico. See id. at 30. Commerce drew distinctions, however, regarding the specificity of the Romanian and Mexican data to the size and price of the raw garlic input.
As to size, Commerce found that the record lacked substantial evidence to conclude that "Mexic[an] garlic bulbs are identical or more comparable to" Chinese garlic bulbs. Id. at 30. Commerce explained that most of Xinboda's evidence on Mexican garlic was "completely unintelligible," and the intelligible articles, or parts thereof, failed to support Xinboda's arguments regarding the comparability of Mexican garlic to Chinese garlic. Id. at 21-22.13
*1301In contrast, Commerce found that substantial evidence supported a finding that Romanian garlic bulbs are "similar in size to the input garlic bulbs." Id. at 30 & n.126 (quoting I & D Mem. at 10).
As to price, Commerce first considered whether Mexican or Romanian data reflected the level of trade at which Xinboda's processor, Excelink, purchased raw garlic. See id. at 26-27. Reiterating its finding in the underlying review that Excelink did not pay "farmgate" prices, Commerce found that Excelink's prices included costs for processing the garlic,14 off-site storage during the non-harvesting months, and transportation from the farmer to Excelink in the harvest season or from rented storage to Excelink in the non-harvesting months. Id. at 26-27. Accordingly, Commerce concluded that Romanian wholesale pricing data better reflected Xinboda's purchasing experience than Mexican farmgate prices. Id. at 27, 30. Commerce also rejected arguments that Romanian garlic prices are distorted by tariff quotas imposed by the European Union ("EU"), and that Mexican garlic prices are distorted by phytosanitary measures excluding Chinese garlic from Mexico. Id. at 29, 30-31.
In view of these distinctions, Commerce selected Romania as the primary surrogate country, id. at 31, and used pricing data reported by the National Institute of Statistics Romania ("NISR") to value Xinboda's raw garlic, I & D Mem. at 13; Remand Results at 26-27. Xinboda challenges Commerce's determination that Romania offered the best available information on the record, focusing on Commerce's findings regarding size, purchasing experience, and distortion in Romanian prices. See generally Xinboda's Remand Opp'n.
From the outset, Xinboda misstates and misapplies the court's inquiry. Xinboda insists that substantial evidence supports the use of Mexican data. See Xinboda's Remand Opp'n at 1-18, 21-29, 31-32. The court's inquiry, however, is whether substantial evidence supports Commerce's use of Romanian data; if so, such use will not be precluded even if substantial evidence also supports the use of Mexican data. See 19 U.S.C. § 1516a(b)(1)(B)(i) ; Matsushita Elec. Indus. Co. v. United States , 750 F.2d 927, 933, 936 (Fed. Cir. 1984) (the possibility of drawing two inconsistent conclusions from the evidence does not preclude the agency's finding from being supported by substantial evidence). Although Commerce's inquiry is directed to which country offers the best available information, the court's inquiry is not the same. Instead, the court considers whether substantial evidence supports the agency's determination. See Zhejiang DunAn Hetian Metal Co., Ltd. v. United States, 652 F.3d 1333, 1341 (Fed. Cir. 2011) (the court is "not to evaluate whether the information Commerce used was the best available, but [ ] whether a reasonable mind could conclude that Commerce chose the best available information"). In so doing, the court is conscious of its role not to "reweigh the evidence or [ ] reconsider questions of fact anew." Downhole Pipe & Equip., L.P. v. United States , 776 F.3d 1369, 1376-77 (Fed. Cir. 2015) (quoting Trent Tube Div., Crucible Materials Corp. v. Avesta Sandvik Tube AB , 975 F.2d 807, 815 (Fed. Cir. 1992) ). Thus, if Commerce's determination is "reasonable given the circumstances presented by the *1302whole record," it will be sustained. Changzhou Hawd Flooring Co., Ltd. v. United States , 42 CIT ----, ----, 324 F.Supp.3d 1317, 1321 (2018) (internal quotation marks and citations omitted). The court addresses Xinboda's specific contentions with the standard of review firmly in mind.
A. The Comparability of Mexican Garlic to Chinese Garlic
Xinboda first contends that Commerce unreasonably accorded more weight to Petitioners' articles discussing Romanian garlic than it did to Xinboda's articles discussing Mexican garlic; the record contained more information establishing the size of Mexican garlic bulbs than it did establishing the size of Romanian garlic bulbs; and Commerce should have informed Xinboda that certain articles it previously accepted were unintelligible or insufficiently translated. Xinboda's Remand Opp'n at 1-4. Defendant and Defendant-Intervenors contend that Commerce relied on the Romanian articles because Xinboda's articles were unintelligible, insufficiently translated, or otherwise failed to establish that Mexican garlic bulbs are more comparable to Chinese garlic bulbs. Def.'s Remand Reply at 10-11; Def.-Ints.' Remand Reply at 7.
As noted, the burden of creating an adequate record before Commerce lies with interested parties. See QVD Food Co. , 658 F.3d at 1324. Although Commerce's regulations require parties to submit English translations of documents written in a foreign language, see Remand Results at 21 & n.100 (citing 19 C.F.R. § 351.303(e) ), most of the information Xinboda submitted was either in Spanish or poorly translated using an online translation generator, see id. at 21-22 (explaining that, of the 68 pages submitted in the underlying administrative review and the additional 214 pages submitted on the record of the remand proceeding, just 16 pages were in English or legibly translated); Xinboda's Remand Opp'n at 3 (acknowledging that its use of an online translation generator resulted in overlapping text and mistranslated words).15 The two fully legible articles consisted of: (1) J.Z. Castellanos, et al., Garlic Productivity and Profitability as Affected by Seed Clove Size, Planting Density and Planting Method , 39 HortScience 1272 (2004) ("Garlic Productivity and Profitability "), see Remand Results at 21 n.99 (citing Xinboda's SV Submission, Ex. SV-6, ECF pp. 167-172);16 Xinboda's Remand Rebuttal, Ex. 3, ECF pp. 289-194);17
*1303and (2) MC Macias Luis Martin Valdez et al., Guide to Cultivate Garlic in Aguascalientes: Producers Brochure No. 21 ("Garlic in Aguascalientes "), see Remand Results at 21 n.99 (citing Xinboda's Remand Rebuttal, Ex. 3, ECF pp.295-302).18
The deficient translations notwithstanding, Commerce addressed Xinboda's core arguments regarding the comparability of Mexican garlic to the subject merchandise. See Remand Results at 18 (referencing Xinboda's arguments regarding the relationship between Mexican commercial garlic classifications and bulb size); id. at 22-23 (addressing the arguments). Commerce first explained that the "main table" discussing garlic size specifications "that Xinboda refer[red] to [is] unintelligible," and appears to reference an international standard. Id. at 22 & n.104 (citing Xinboda's Remand Rebuttal, Ex. 3, ECF p. 493).19 Although the translated table is partially legible, see Xinboda's Remand Rebuttal, Ex. 3, ECF p. 493, it is contained in an article replete with translation issues, see id. , ECF pp. 486-498. More important, as Commerce noted, the table's reference to international standards for garlic classification obviates its relevance to the nature of Mexican garlic. See id. , ECF p. 486.20 Even if the table reflected Mexican garlic sizes, Xinboda has pointed to no record evidence to support its theory that the number of size categories available are correlated to production volumes. See Xinboda's Remand Opp'n at 5-6 (asserting that the existence of several categories above size 3 (i.e., sizes 4 to 12), applicable to garlic sized 35 millimeters ("mm") or more), with just size 3 applicable to garlic sized less than 35 mm, demonstrates large bulb garlic production).
Commerce further explained that information regarding commercial classifications for garlic contained in Garlic Productivity and Profitability referred solely to the Tacatzcaro variety, and the record lacked evidence suggesting that Tacatzcaro garlic was sold in Mexico during the period of review or was "the main type of garlic produced and sold in Mexico." Remand Results at 22 & n.105 (citing Xinboda's Remand Rebuttal, Ex. 3, ECF p. 289). Xinboda asserts that the class numbers, commercial classifications, and bulb diameters nevertheless reflect Mexican garlic standards generally and "show that the majority of garlic grown and traded commercially *1304is large bulb garlic." Xinboda's Remand Opp'n at 7. The relevant table lists seven categories of Tacatzcaro garlic,21 ranging from 5 (bulb diameter of 40-45mm) to 11 (bulb diameter greater than 70mm). See Xinboda's Remand Rebuttal, Ex. 3, ECF p. 289 (Table 1). At most, the table suggests that Tacatzcaro garlic is generally large bulb, but, as Commerce noted, the table offered no indication regarding the volume of Tacatzcaro garlic sold in Mexico or the amount of Tacatzcaro garlic produced and sold in relation to Mexico's total garlic production. See Remand Results at 22.
In contrast to Xinboda's evidence purporting to establish the size of Mexican garlic, the information Commerce relied upon to substantiate the size of Romanian garlic is legible and fully translated. See id. at 23 & n.110 (citing I & D Mem. at 10); Pet'rs' Rebuttal Comments on Surrogate Country Selection (June 11, 2015) ("Pet'rs' Rebuttal SC Comments"), Ex. ROM-1, PR 191-94, PRJA Tab 5, Suppl. CRJA Tab 3, Suppl. PRJA Tab 3.22 That the record contained more pages ostensibly about some aspect of Mexican garlic than Romanian garlic is of no moment if Commerce cannot read those pages, or most of the content therein.23
Xinboda's reliance on 19 U.S.C. § 1677m(d) also fails. Section 1677m(d) states the procedures Commerce must undertake in connection with deficient "response[s] to a request for information." 19 U.S.C. § 1677m(d).24 Read in context, however, *1305section 1677m(d) applies to a respondent's questionnaire responses, which are distinct from the voluntary submission of surrogate value data. See Qingdao Sea-Line Trading Co., Ltd. v. United States , Slip Op. 12-39, 2012 WL 990904, at *7 n.9 (CIT Mar. 21, 2012).25 Providing Xinboda with additional time to translate the articles would have amounted to an extension of time for the submission of factual information that should have been translated in the first instance. See 19 C.F.R. § 351.303(e).
Xinboda next contends that Mexico's exports of fresh garlic to the United States speaks to Mexico's production of large bulb garlic and cites several pieces of evidence in support thereof. Xinboda's Remand Opp'n at 13-15. The Government dismisses this argument, as Commerce did, on the basis that production levels and export amounts are relevant to the significant producer analysis and, even then, the significance of production is not judged against China's production levels. Def.'s Remand Reply at 14; Remand Results at 23. Defendant-Intervenors contend that the size of exported garlic bulbs "is not representative of the size of bulbs that are typically grown and harvested in Mexico." Def.-Ints.' Remand Reply at 9.
It is Commerce's province to weigh the evidence in the first instance, not the court's. See POSCO v. United States , 42 CIT ----, ----, 296 F.Supp.3d 1320, 1349 (2018) (citing Bowman Transp., Inc. v. Ark.-Best Freight System, Inc. , 419 U.S. 281, 285-86, 95 S.Ct. 438, 42 L.Ed.2d 447 (1974) ). In view of Commerce's conclusory response to this argument, the court's inquiry is whether the evidence upon which Xinboda relies undermines Commerce's determination to render it unsupported by substantial evidence. See Nippon Steel Corp. v. United States, 337 F.3d 1373, 1379 (Fed. Cir. 2003) (the court's review must account for "the record as a whole, including evidence that supports as well as evidence that fairly detracts from the substantiality of the evidence") (internal quotation marks and citation omitted).26 The *1306court concludes that Commerce's determination is supported by substantial evidence.
Xinboda first points to an excerpted portion of a table contained in an article titled "Garlic Production [Chain]." Xinboda's Remand Opp'n at 13 (citing Xinboda's Remand Rebuttal, Ex. 3, ECF p. 436 (Table 19) ); see also Xinboda's Remand Rebuttal, Ex. 3, ECF pp. 418-50 (Manuel E. Well Spinosa et al., Foundation Produces [sic ] Querétaro, Garlic Production Chain) ("Garlic Production Chain "). Although the table itself is legible, Garlic Production Chain as a whole is replete with mistranslations and other errors, see Xinboda's Remand Rebuttal, Ex. 3, ECF pp. 418-50. Table 19 lists the sizes of Mexican garlic exported to the United States and Canada from 1999 to 2002, predating the period of review by roughly 11 years. See Xinboda's Remand Rebuttal, Ex. 3, ECF p.436; Final Results , 81 Fed. Reg. at 39,897. Xinboda excerpted only part of the table, however, listing Mexican exports of garlic sized giant, extra big, big, extra jumbo, jumbo, and super jumbo. See Xinboda's Remand Opp'n at 13. Xinboda omitted the part of the table listing Mexican exports of garlic sized unspecified, super colossal, little, colossal, and medium. See Xinboda's Remand Rebuttal, Ex. 3, ECF p. 436. The table does not state the respective amounts of each garlic size exported. See id. Thus, at most, the table indicates that, more than a decade ago, Mexico may have exported garlic in a range of sizes to the United States and Canada.27
Xinboda's subsequent argument that Mexico is the largest supplier of garlic to the United States after China is unsupported by the cited evidence. See Xinboda's Remand Opp'n at 15 (citing Xinboda's SC Comments, Ex. 2). Xinboda's argument regarding Mexico's supply of fresh garlic to the United States "when the Chinese and domestic crop is out of season" is inapposite to the inquiry into Mexican garlic production generally for purposes of assessing the quality of Mexican surrogate value data.28 See Xinboda's Remand Opp'n at 15-16 (citing Pet'rs' Remand Rebuttal, Attach. ITC-1 at I-11). Thus, Xinboda's appeal to Mexican export data is unavailing.
*1307Xinboda's final contention concerns Petitioners' arguments to Commerce regarding dissimilarities between Mexico's and China's respective climates and similarities between Romania's and China's respective climates. Xinboda's Remand Opp'n at 16-18. Defendant and Defendant-Intervenors do not substantively respond to Xinboda's climate-related discussion. See Def.'s Remand Reply at 11, 14-15; Def.-Ints.' Remand Reply at 6-10.
Commerce, for its part, noted Petitioners' arguments but did not rely on them. See Remand Results at 15. Xinboda's reliance on certain record evidence to draw connections between Chinese and Mexican garlic growing climates nevertheless fails to undermine Commerce's determination. See Nippon Steel , 337 F.3d at 1379 (directing the court to "review the record as a whole"). Xinboda fails to provide support for its asserted linkage between garlic growing and the presence of a cold semi-arid climate. See Xinboda's Remand Opp'n at 18. Xinboda further fails to demonstrate that the areas in Mexico exhibiting a cold semi-arid climate are the major garlic growing regions. Compare Pet'rs' Remand Rebuttal, Attach. Climate-4 (listing Mexican regions with cold semi-arid climates), with Xinboda's Remand Opp'n at 17 (listing major garlic growing regions of Mexico) (citing Xinboda's Remand Rebuttal, Ex. 2).
Xinboda's arguments regarding the comparability of Mexican garlic to the subject merchandise do not undermine Commerce's reliance on Romanian data because Commerce addressed these arguments to the extent that they were relevant and supported by legible record evidence, and Xinboda's arguments do not detract from the substantial evidence concerning Romanian garlic size on which Commerce relied.
B. The Comparability of Romanian Garlic to Chinese Garlic
Xinboda challenges Commerce's reliance on the weight of Romanian garlic as set forth in Growing Vegetables to draw conclusions about the Romanian garlic size. See Xinboda's Remand Opp'n at 18. Xinboda contends that Commerce impermissibly relied on a 2012 online advertisement from a Chinese exporter to establish a weight-to-size ratio because the advertisement is outdated and "from a little known exporter." Id. at 19. Xinboda further contends that Growing Vegetables relies on 1998 data and fails to establish country-wide production of large bulb garlic. See id. at 20. Xinboda also contends that Romanian yield size suggests that Romanian garlic is generally not large bulb. Id. at 20.
Defendant contends that Commerce's choice between "two imperfect options"-poorly translated or illegible Mexican data and a "single, dated article" requiring a "straightforward inference with respect to a correlated ratio"-is left to the agency's discretion. Def.'s Remand Reply at 10 (citation omitted). Defendant-Intervenors contend that the "conversion of bulb weights ... to diameters required only a simple calculation," and Xinboda points to no evidence contradicting Commerce's conversion. Def.-Ints.' Remand Reply at 10-11.
The Growing Vegetables article upon which Commerce relied to substantiate the size of Romanian garlic contains a table listing the types of garlic cultivated in Romania. See Remand Results at 30 & n.126 (citing I & D Mem. at 10); I & D Mem. at 10 & n.52 (citing Pet'rs' Rebuttal SC Comments at 10); Pet'rs' Rebuttal SC Comments at 10 (citing Pet'rs' Rebuttal SC Comments, Attach. ROM-1). In the spring, those varieties are medium-sized and weigh 20 to 30 grams. See Pet'rs' Rebuttal SC Comments, Attach. ROM-1, ECF p. 118. In the fall, Romanian garlic *1308consists of one medium-sized variety weighing 25 to 35 grams, and two large-sized varieties weighing 40 to 50 grams and 40 to 60 grams, respectively. See id.
To convert weight to size, Commerce used an advertisement from an online marketplace regarding the sale of a 250-gram bag of fresh garlic consisting of four bulbs with 60-centimeter ("cm") diameters. See Remand Results at 23-24; Pet'rs' Rebuttal SC Comments, Attach. PRC-1, ECF p. 153. As Commerce explained, the advertisement can "be used to calculate a general weight ratio[:] ... 250 grams/4 bulbs = 62.5 grams per bulb," Remand Results at 24; i.e., about one gram per centimeter of the diameter. Absent alternative information on the record, the advertisement was the best available for Commerce to use and Xinboda points to nothing to suggest the conversion is unreasonable.
Xinboda's arguments regarding Romanian garlic size and yield size also fail. Chinese garlic ranges in size from about 1 ½ to 2 ½ inches in diameter, See Decision Mem. for the Prelim. Results of the 2013-2014 Antidumping Duty Admin. Review at 22 & nn.137-38 (citations omitted), PR 398, CJA Tab 51, PJA Tab 51, the equivalent of about 3.81cm to 6.35cm in diameter. Applying the weight to size ratio of one gram per centimeter of diameter shows that two of the fall garlic varieties are within the size range of the subject merchandise. See Pet'rs' Rebuttal SC Comments, Attach. ROM-1, ECF p. 118. As to yield, record evidence demonstrates that Romanian garlic output can range from 5-6 metric tons ("MT") per hectare to 10-12 MT per hectare, indicating the potential for Romanian garlic to be weighted to the larger varieties in the Growing Vegetables table. See id. , Attach. ROM-1, ECF p. 120. In sum, Commerce found evidence of large bulb garlic production in Romania, and the record supports that finding.
C. Level of Trade
Xinboda contends that Commerce erred in relying on Romanian wholesale prices to value its garlic because Xinboda purchases garlic on a farmgate basis, and the Mexican pricing data consists of what Xinboda considers to be farmgate prices. See Xinboda's Remand Opp'n at 21-29. Xinboda further contends that Commerce's selection of a wholesale price is inconsistent with the agency's practice in prior administrative reviews. Id. at 24. Defendant and Defendant-Intervenors contend that this record supports Commerce's determination that Xinboda's garlic price contains post-harvesting charges that are not indicative of farmgate prices. Def.'s Remand Reply at 16-18; Def.-Ints.' Remand Reply at 12-15.
Commerce defined the term "farmgate" as "the purchase price of raw garlic as it is harvested with no further processing or handling, and including no additional charges .... [I]t is garlic, immediately following harvest, that has not been sorted, cleaned, processed, stored, transported or in any other way handled or modified." Remand Results at 26 & n.115 (citation omitted). Commerce examined the input for which it was seeking a surrogate value and determined that Xinboda's processor, Excelink, was not making farmgate purchases as Commerce defined the term. See id. at 26-27 & nn.118-19 (discussing evidence of post-harvest processing, rented cold-storage, and transportation) (citing Suppl. Questionnaire Resp.-Shenzen Xinboda (Sept. 1, 2015) at 4, CR 137-45, PR 304-05, PRJA Tab 7, Suppl. CRJA Tab 4, Suppl. PRJA Tab 4; Second Suppl. Questionnaire Resp.-Shenzhen Xinboda (Nov. 3, 2015) at 3-5, CR 163, PR 365, PRJA Tab 8, Suppl. CRJA Tab 5, Suppl. PRJA Tab 5). Taking this evidence into consideration along with Commerce's use of the intermediate *1309input methodology, Commerce determined that the Romanian pricing data was more comparable than Mexican pricing data. Remand Results at 27; see also I & D Mem. at 28-29 (discussing the intermediate input methodology by which Commerce seeks a surrogate value for the price of fresh garlic rather than the factors of production used in producing fresh garlic). Commerce noted that in selecting surrogate values it need not "match the respondent's exact production experience." Remand Results at 27 & n.120 (citing I & D Mem. at 16; Nation Ford Chem. Co. v. United States , 166 F.3d 1373, 1377 (Fed. Cir. 1999) ). In other words, the surrogate value need not be perfect. Here, while neither the farmgate prices, as Commerce accepted the Mexican pricing data to be, nor the wholesale prices, as Commerce understood the Romanian data to be, may precisely describe the stage of distribution at which Excelink purchases garlic, Commerce's decision to select Romanian wholesale pricing data, which includes certain post-harvesting and storage costs, is supported by substantial evidence. Xinboda's arguments to the contrary are unavailing.
To support the assertion that Excelink purchased garlic on a farmgate basis, Xinboda relies on a different understanding of the term that allows for the inclusion of some processing or storage costs. See Xinboda's Remand Opp'n at 27 (citations omitted). Commerce's definition, however, is reasonable and consistent with the definition Commerce used in prior administrative reviews. See Issues and Decision Mem. for the Final Results and Rescission, in Part, of the Antidumping Duty Admin. Review of Fresh Garlic from the People's Republic of China, A-570-831 (June 10, 2013) ("AR 17 Decision Mem.") at 14 n.56, available at https://enforcement.trade.gov/frn/summary/prc/2013-14329-1.pdf (last visited Dec. 18, 2018); Issues and Decision Mem. for Fresh Garlic from the People's Republic of China: Final Results of the 2009-2010 Admin. Review, A-570-831 (June 4, 2012) ("AR 16 Decision Mem.") at 19-20, available at https://enforcement.trade.gov/frn/summary/prc/2012-14152-1.pdf (last visited Dec. 18, 2018).
Specifically, Xinboda points to Commerce's successful defense of its selection of farmgate prices in the 17th administrate review as "the law of the case" requiring Commerce to use farmgate prices in this review. Xinboda's Remand Opp'n at 25. However, the law of the case doctrine applies to issues that have previously been resolved. See Intergraph Corp. v. Intel Corp. , 253 F.3d 695, 697 (Fed. Cir. 2001) ("[L]aw of the case doctrine 'expresses the practice of courts generally to refuse to reopen what has been decided.' ") (quoting Messenger v. Anderson, 225 U.S. 436, 444, 32 S.Ct. 739, 56 L.Ed. 1152 (1912) ). Commerce's selection of pricing data to value Xinboda's raw garlic purchases in this period of review has not previously been resolved; thus, the law of the case doctrine is inapposite. Further, Commerce's decision in the 17th administrate review illustrates why Commerce may reach different conclusions in separate administrative reviews based on different record facts. See generally Jiaxing Brother Fastener Co., Ltd. , 822 F.3d at 1299 ("each administrative review is a separate exercise of Commerce's authority that allows for different conclusions based on different facts in the record") (citation omitted). In the 17th administrative review, while Commerce selected a farmgate-approximate price over a price resembling a wholesale price, the agency similarly discussed the fact that Xinboda's garlic was not obtained at farmgate prices, but, as here, included some post-harvest processing. Taking that into consideration, in relation to the surrogate values placed on that record (including the Azadpur sales prices from India, which are *1310not at issue in this review) Commerce made its record-specific determination. AR 17 Decision Mem. at 14-23.
There is one aspect of Commerce's determination, however, that requires further consideration by the agency. In the underlying proceeding, Commerce added transportation costs to the Romanian wholesale prices to account for delivery of the raw garlic "from the farmer or storage facility to Excelink." See I & D Mem. at 29 & n.174 (citation omitted). Information placed on the record of the remand proceeding suggests that NISR prices "likely include any transportation costs , storage costs, processing or packaging costs, and a profit mark-up." Xinboda Remand Rebuttal, Ex. 4, ECF p. 523 (emphasis added). Thus, while the court sustains Commerce's determination that Romanian prices are more comparable to Excelink's purchasing experience, a remand is required for Commerce to clarify its justification for adding transportation costs to the wholesale prices.
D. Distortion in the Romanian Garlic Market
Xinboda contends that a garlic tariff quota imposed on Chinese garlic imported into Romania after Romania joined the EU in 2007 has distorted Romanian garlic prices. Xinboda's Remand Opp'n at 29-30. Defendant and Defendant-Intervenors contend that Commerce correctly declined to infer that the price increases after Romania joined the EU in 2007 resulted either from a tariff quota (as Xinboda had asserted) or from "the removal of unfairly traded garlic imports" (as Petitioners had asserted). Def.'s Remand Reply at 18-19; Def.-Ints.' Remand Reply at 15-16.
On this issue, Commerce explained that the record failed to demonstrate that EU-imposed tariff quotas distorted or otherwise increased Romanian garlic prices. I & D Mem. at 14; see also Remand Results at 29 & n.124 (citing I & D Mem. at 11-14). According to Commerce, at most, Xinboda had demonstrated "a temporal correlation" between Romania's accession to the EU and increased prices, not causation. I & D Mem. at 14. Commerce faulted Xinboda for "presum[ing] that such a relationship exist[ed]" absent evidence "quantifying such a relationship." Id.
Before the court, Xinboda surmises that "[t]he tariff quota is the only reasonable explanation for the sustained price increase." Xinboda's Remand Opp'n at 30. Upon review of the entirety of the record, substantial evidence supports Commerce's decision not to infer the existence of such a relationship. See Nippon Steel , 337 F.3d at 1379.
As of April 1, 2007, the EU imposed tariff quotas on imports of fresh garlic from China, Argentina, and other third countries. See Rebuttal Final Surrogate Value Submission (Nov. 12, 2015), Ex. SV-2, ECF pp. 419, 426, PR 371, CJA Tab 42, PJA Tab 42. Pursuant thereto, the first 33,700 MT of garlic imported into the EU from China was subject only to a 9.6 percent ad valorem duty; thereafter, any additional garlic was subject to the 9.6 percent ad valorem duty plus an additional duty of 1,200 Euros per MT. See id. , ECF pp. 419, 425-26 (further providing for tariff quotas of 19,147 MT on garlic imports from Argentina and 6,023 MT on garlic imports from other third countries). In 2014, the tariff quota applicable to China increased to 46,075 MT, while the tariff quotas applicable to Argentina and other third countries remained the same, thereby increasing the quantity of garlic that could be imported into the EU at 9.6 percent duties to more than 70,000 MT. Id. , ECF p. 436. In 2007, EU countries collectively imported between 60,000 and 80,000 *1311MT of fresh garlic from other countries, with between 30,000 and 40,000 MT originating in China. Id. , ECF p. 419. However, Xinboda provided no evidence indicating that the volume of fresh garlic imported into the EU increased between 2007 and 2014; thus, it is speculative to assume that the above-quota duty had any effect on import quantities, let alone pricing within the EU generally, or Romania specifically, during the period of review.
In light of record evidence indicating that most, if not all, of the fresh garlic imported into the EU was not subject to the additional duty of 1,200 Euros, substantial evidence supports Commerce's decision not to infer a causal relationship between EU-imposed tariff quotas and increased Romanian garlic prices. See Remand Results at 29; I & D Mem. at 14.
In sum, Commerce's determination that Romanian data better fulfilled the contemporaneity and product-specificity criteria of its factors data analysis is supported by substantial evidence. Accordingly, notwithstanding the need to address the addition of transportation costs to Romanian pricing date, Commerce's selection of Romania as the primary surrogate country is sustained.
II. Movement Expenses
For the Final Results , Commerce relied on information contained in the World Bank's Doing Business 2015: Romania ("Doing Business Romania ") report to calculate a surrogate value for brokerage and handling and inland truck freight pursuant to 19 U.S.C. § 1677a(c)(2)(A). I & D Mem. at 30-31; see also Pet'rs' Surrogate Value Comments (June 17, 2015) ("Pet'rs' SV Comments"), Ex. 6, PR 275-80, CJA Tab 26, PJA Tab 26 (copy of the report). Specifically, Commerce "calculated a per kilogram brokerage and handling and inland freight [cost] using the price data to export standardized cargo of ten metric tons in a standard 20-foot container as published in Doing Business Romania ." I & D Mem. at 31. In other words, Commerce used the costs reported in Doing Business Romania as the numerator in the equation with a payload weight of 10,000kg (ten MT) as the denominator. See Surrogate Values for the Prelim. Results (Nov. 30, 2015) ("Prelim. SV Mem."), Exs. 8, 9, PR 401-02, CJA Tab 53, PJA Tab 53. The calculation of inland freight required additional inquiry into the distance traveled (i.e., cost of inland transportation and handling per kilogram per kilometer). See id. , Ex. 8. According to Commerce, "the container payload weight of 10,000kg [is] explicitly stated in the Doing Business methodology," and "is one of the assumptions in all Doing Business reports." I & D Mem. at 30 & n.184 (citing Pet'rs' SV Comments, Ex. 6).
Xinboda challenges Commerce's use of the 10,000kg denominator. See Xinboda's 56.2 Br. at 32-35. Though apparently accepting the 10,000kg denominator weight as an assumption of the report, Xinboda contends that "does not infer that the ... weight of the container is the basis of the cost." Id. at 32. According to Xinboda, record evidence demonstrates that brokerage and handling fees "are based on an entire container and not the weight of its contents." Id. at 33. Xinboda contends that Commerce should, therefore, replace the 10,000kg denominator with "the maximum weight of a container or the average weight of Xinboda's containers." Id. at 35; see also Pl. Shenzhen Xinboda Indus. Co., Ltd. Reply Br. ("Xinboda's 56.2 Reply") at 18-20, ECF No. 53.
Defendant contends that Commerce's calculations are correct because Doing Business Romania provides data relevant to a "standard shipment of goods," which is "a dry-cargo, 20-foot-full container load"
*1312that is "assumed to weigh 10,000kg." Confidential Def.'s Resp. to Consol. Pls.' Mots. for J. Upon the Agency R. ("Def.'s 56.2 Resp.") at 53-54, ECF No. 47 (citing Pet'rs' SV Comments, Ex. 6 at 2, 69). Defendant-Intervenors similarly contend that because "World Bank survey [respondents] are asked to assume that the payload weight is 10,000kg, the reported costs are based on this assumption" and using a different denominator "would distort the costs." Confidential Def.-Ints.' Resp. in Opp'n to Pls.' Mots. for J. on the Agency R. ("Def.-Ints.' 56.2 Resp.") at 46, ECF No. 46.
Before addressing Xinboda's arguments for alternative denominators, the court must begin by assessing whether Commerce's decision to apply a 10,000kg denominator is supported by substantial evidence and in accordance with law. It is not.
In the Issues and Decision Memorandum, Commerce quoted extensively from its determination in Certain Nails from the People's Republic of China as support for its use of the 10,000kg denominator. I & D Mem. at 31 & n.187 (quoting Certain Nails from the People's Republic of China , 78 Fed. Reg. 16,651 (Dep't Commerce Mar. 18, 2013) (final results of third antidumping duty admin. review; 2010-2011). The decision memorandum accompanying that determination notes that Commerce calculated brokerage and handling using data supplied by a Doing Business report on Thailand. See Certain Steel Nails from the People's Republic of China: Issues and Decision Mem. for the Final Results of the Third Antidumping Duty Admin. Review, A-570-909 (Mar. 5, 2013) at 34, available at https://enforcement.trade.gov/frn/summary/prc/2013-06173-1.pdf (last visited Dec. 18, 2018). Therein, Commerce explained that it calculated brokerage and handling "by dividing the total charge by 10 tons which is found under 'Trading Across Borders Methodology: Assumptions about the Business,' " which states, "The traded product travels in a dry cargo, 20 foot, full container load. It weighs 10 tons ." Id. at 34 (emphasis added).
Here, however, the Doing Business Romania report Commerce relied upon merely assumes the presence of "a dry-cargo, 20-foot-full container load," Pet'rs' SV Comments, Ex. 6 at 69 (emphasis added); it does not provide any assumption regarding the container's weight, see id. Commerce's assertion that the 10,000kg payload weight is "explicitly stated in the Doing Business methodology" is, simply, incorrect, I & D Mem. at 30 (citing Pet'rs' SV Comments, Ex. 6) (emphasis added), and Commerce's assertion that a 10,000kg payload weight "is one of the assumptions in all Doing Business reports" is unsupported by the record, see id. (emphasis added).
Shipping information supplied by Xinboda suggests that weight or volume is taken into consideration only when shipping less than full container loads. See Prelim. Surrogate Values Submission (June 17, 2015), Ex. SV-21, ECF pp.60, 70-72, PR 281-85, CJA Tab 27, PJA Tab 27 (international freight forwarders offering rates based solely on the size of the container irrespective of weight); id. , Ex. SV-21, ECF pp.61-66 (distinguishing between full container loads (which are based on container size) and less than full container loads (which are based on weight or volume) ).29 While the record therefore supports the inference that the costs reported in Doing *1313Business Romania reflect the cost of shipping full container loads, the record does not support the inference that those costs are based on a 10,000kg payload weight. See I & D Mem. at 30.30 Although Commerce must identify a suitable denominator to calculate Xinboda's movement expenses on a per unit basis, without more, Commerce's use of a 10,000kg denominator is arbitrary.31 See Changzhou Wujin Fine Chemical Factory Co., Ltd. v. United States , 701 F.3d 1367, 1377 (Fed. Cir. 2012) (providing for review of Commerce's reasoning pursuant to "the arbitrary and capricious (or contrary to law) standard") (citing Motor Vehicle Mfrs. Ass'n v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 48-49, 103 S.Ct. 2856, 77 L.Ed.2d 443 (1983) ).32 Accordingly, this issue is remanded to the agency for further consideration.33
CONCLUSION & ORDER
In accordance with the foregoing, it is hereby
ORDERED that Commerce's Remand Results are sustained with respect to Commerce's selection of Romania as the primary surrogate country and selection of Romanian pricing data as the surrogate value for raw garlic, as set forth in Discussion Section I; and it is further
ORDERED that Commerce's Remand Results are remanded with respect to Commerce's addition of transportation costs to the surrogate value for raw garlic, as set forth in Discussion Section I.C; and it is further
ORDERED that Commerce's Final Results are remanded with respect to Commerce's calculation of Xinboda's movement expenses as set forth in Discussion Section II; and it is further
ORDERED that, in the event Commerce amends the antidumping margin assigned to Xinboda, Commerce reconsider the separate rate assigned to non-mandatory respondents; and it is further
ORDERED that Commerce shall file its remand results on or before April 2, 2019; and it is further *1314ORDERED that the deadlines provided in USCIT Rule 56.2(h) shall govern thereafter; and it is further
ORDERED that any opposition or supportive comments must not exceed 5,000 words.