Traditionally, PMF was done over a short period of time in case of intense measurement campaigns or seasonal-wise for long-term datasets. Although seasonal PMF already provides better PMF results that performing PMF on a year-long dataset, the problem of static factor profiles over a defined time period remains. This is especially visible between seasons or special events. With rolling PMF, this is not an issue anymore.
Across ambient datasets investigated, rolling PMF performed better than seasonal PMF in terms of correlation with external measurements, especially in between-season periods. Compared to the synthetic dataset, this multi-site comparison also showed rolling-seasonal similarity and moderate improvement in rolling PMF. Find more details on the comparison of rolling vs seasonal PMF here.
Rolling vs seasonal PMF: M. Via et al., 2022, https://doi.org/10.5194/egusphere-2022-269
Rolling PMF in general: F. Canonaco et al., 2021, https://doi.org/10.5194/amt-14-923-2021