Save both original pages before choosing a side
A relative sends a calendar screenshot, but another website gives a different date or seasonal label. Start by preserving the complete sources: publisher, URL, displayed year, access date and legend. A single cropped cell is often insufficient to explain a discrepancy.
Describe the problem precisely: “These two tables place the same solar term on different Gregorian dates in the same year.” Avoid beginning with “This table is wrong,” or assuming every difference comes from a time zone.
First check the year and the kind of field
Compare the Gregorian year, any lunar-year label, the month and any leap-month notation. Then ask whether the fields serve the same purpose. One may display a Gregorian date and the other a lunar day; one may mark a solar term while another carries a customary annotation. Different kinds of information cannot be compared as though they were identical.
Use four lines: wording in source A, wording in source B, each field’s heading, and the single item you intend to compare. Transcribe the original first. Immediate reformatting can accidentally erase the difference you are trying to understand.
Next inspect the legend and the layout
Check whether the table is read across rows or down columns and what colours, underlining or month labels mean. A PDF converted into a phone image can make text appear to belong to a neighbouring day. Extracted text may also lose the relationships established by grid lines.
Circle the relevant position and adjacent cells on each original. Keep the title and legend with the main table, as when reading the Observatory’s annual conversion PDFs. Search-result snippets are not a substitute for inspecting the table itself.
Check time standards when the disputed information involves time
If the discrepancy concerns the time or date of a solar term, look for each source’s stated time standard and location. The Hong Kong Observatory explains Hong Kong Standard Time as UTC plus eight hours. For another region, establish the standard actually used before converting; the language of a website does not establish its time zone.
If one timestamp has no stated standard, record “time standard unconfirmed.” Do not use the numerical difference to prove which time zone its publisher must have chosen.
Return to the publisher’s version and notes
Find the publication page. Check whether a screenshot is old, whether its link opens the intended year, and whether corrections exist. The Observatory’s conversion-table page includes notes about distant-future lunar-phase and solar-term calculations near midnight. Read such limitations, but do not treat their existence as an explanation for every apparent discrepancy.
Record what you actually found: “Source A includes this note; no corresponding explanation has been located in source B.” When contacting a publisher, provide the year, field names, original links and marked positions rather than an unsupported conclusion.
Use a discrepancy record another person can follow
A practical example is: “Item: date of a solar term. Source A: year confirmed; time standard stated. Source B: same year, but the screenshot omits its legend and time information. Completed: located B’s original page. Outstanding: confirm whether its label is a date or an event time. Current conclusion: insufficient information to identify the cause.”
This is a working format, not a claim that an actual calendar is incorrect. It shows a helper what has already been checked and what evidence is still needed.
Keep the result and make arrangements deliberately
Once the issue is clarified, record the checking date, original pages and explanation. Retain both sources if it remains unresolved. A family gathering can still be discussed on practical grounds while a cultural detail is being checked.
For help with the labels themselves, read our Chinese calendar glossary. If you are considering a Feng Shui and BaZi course, bring the complete comparison and ask which sources and calculation conventions it uses.