Caveats in using the WRIT pages
Users should take into account the issues below when using the WRIT webpages. Please use the reanalyses.org WRIT wiki page to make comments or suggestions.
- WRIT time-series and trajectory tools use the closest grid-point(s) to the user selected latitude and longitude. When grid resolutions differ, the exact location of the box used will not be the same and some dataset differences may be due to the fact that the locations are not identical.
- While the general circulation models used in the reanalysis systems remain unchanged throughout the time period of the reanalysis dataset (with the exception of CFSR), observational inputs to the systems are not. Interpretations of differences between reanalyses will need to take this into account. The introduction and variations of satellite data creates inhomogeneities in some reanalysis fields. Please research each dataset to see how these inputs may have changed over time.
- Areas with a high density of observations available for assimilation into the reanalysis fields are more likely to have a smaller error and be more consistent with independent observations. Areas with a low density of observations can be expected to have higher error.1
- When using MERRA and MERRA-2 data, be aware that the model outputs missing values below the surface and does not interpolate or otherwise fill in those values, which is something all of the other reanalysis models do. This will particularly impact the near-surface pressure levels over higher terrain. Maps may have large areas of missing values when differencing datasets and time series will be biased in MERRA towards ocean and flatter land surfaces. Plotting the surface pressure for the dates you have selected will give a better idea of where this happens.
- Also, see the general description of reanalysis at the Climate Data Guide at NCAR.
- Most of the anomalies are calculated using a 1981-2010 climatology time period. Some of the datasets on the page are anomaly datasets with a different climatology time range. The time period of the climatology will impact anomaly values. Note you can subtract out a different set of years, thus creating your own climatology.
- The model used in calculating the trajectories is unaware of the surface and uses pressure level data only. In all of the models used on the page, data is interpolated or otherwise filled in below the surface. This means that trajectories may go below the surface.
- The reanalysis data used is 4-times daily and not a higher temporal frequency. It is possible the 4 times daily data will not capture the full diurnal cycle at each grid point.