If you have a table named [YourTable] with the data...
id date
------ ----------
510010 2013-09-12
510010 2013-09-10
510010 2013-09-05
510010 2013-09-14
510012 2013-09-14
510012 2013-09-10
...then you can create a saved query named [CalculateDaysAway] in Access...
SELECT
[id],
[date],
Abs(DateDiff("d", [date], Date()) AS DaysAway
FROM YourTable
...which (when run on 2013-09-11) returns
id date DaysAway
------ ---------- --------
510010 2013-09-12 1
510010 2013-09-10 1
510010 2013-09-05 6
510010 2013-09-14 3
510012 2013-09-14 3
510012 2013-09-10 1
Now you can use that query as a basis for the query to return the closest date(s)...
SELECT
CalculateDaysAway.[id],
CalculateDaysAway.[date]
FROM
CalculateDaysAway
INNER JOIN
(
SELECT
[id],
MIN(DaysAway) AS MinOfDaysAway
FROM CalculateDaysAway
GROUP BY [id]
) AS MinDays
ON CalculateDaysAway.[id] = MinDays.[id]
AND CalculateDaysAway.DaysAway = MinDays.MinOfDaysAway
...which (when run on 2013-09-11) returns
id date
------ ----------
510010 2013-09-12
510010 2013-09-10
510012 2013-09-10
与恶龙缠斗过久,自身亦成为恶龙;凝视深渊过久,深渊将回以凝视…