The graph shows 12-month moving averages with the final data point appearing to show the average of July 2009 to June 2010, so in any case it doesn’t show a comparison of 2010 anomalies.
The most likely reason for the discrepancy is that BEST don’t have 2010 data finalised yet so in their comparison setup January to June 2010 monthly anomalies are set to zero, creating a downward diagonal in the moving average.
Also, the 2010 anomaly in UAH is in good agreement with the surface records.