International trade has been considered a critical driving force of material flows and their environmental pressures, which has been a global research hotspot. The United Nations Commodity Trade Statistics Database (UN Comtrade) is the original and probably the most widely-used data source to support the physical trade analysis. However, data discrepancies have been discovered in UN Comtrade, which may lead to diametrically conflicted conclusions if not properly addressed. To promote applications of UN Comtrade, this article reviews data statistics criteria and preprocessing procedures, discusses three main data quality issues (outliers, missing values, and bilateral asymmetries), and reviews methods to explore adequate options. It is revealed that data quality issues existed in data of almost all the commodities, reporters, and periods, but existing methods are subject to certain limitations. Furthermore, this article presents a brief introduction of our following work on addressing these issues.

国际贸易一直被认为是物质流动及其环境压力的关键驱动力，已成为全球研究热点。联合国商品贸易统计数据库 (UN Comtrade) 是支持实物贸易分析的原始数据来源，并且可能是使用最广泛的数据来源。然而，在联合国商品贸易统计中发现了数据差异，如果处理不当，可能会导致完全矛盾的结论。为了促进联合国COMTRADE的应用，本文回顾了数据统计标准和预处理程序，讨论了三个主要数据质量问题（离群值，缺失值和双边不对称），并审查了探索足够选项的方法。据透露，几乎所有商品、记者、时段的数据都存在数据质量问题，但现有的方法受到一定的限制。此外，本文还简要介绍了我们在解决这些问题方面的后续工作。