Analysis of the gaps in the Hydra-Ring Databases

When generating the hydro-dynamic database for the various regions (RMM, IJVD, and Upper Rivers) some information is discarded due to unstable results of the WAQUA simulations at both the dike location and closest axis location (see Deltares, 2016a). This discarded information results in gaps in the database for some locations, i.e. for some combinations of wind, sea/lake water level, and discharge, local conditions are not available. For some locations, there is even no information available in the database at all. For locations with no information available, Hydra-Ring computations cannot be done and an alternative is needed. Where there is information available, but there are gaps, Hydra-Ring computations can be done; however, there is some concern that these gaps would result in unrealistic Hydra-Ring results. Therefore, the effect of these gaps on the results was investigated"for RMM and IJVD dike locations. Following this investigation, new databases are made for both RMM and IJVD. These databases mainly contain locations where there was no significant influence of these gaps. These locations are flagged as accurate. They also contain some of the locations where there is a significant influence of these gaps, and locations for which no local conditions were available at all (excluded from the original databases). However, the locations in the last two classes are added using a different filling technique (using manually selected backup locations). These locations will be flagged as "less" accurate. In the end 880 locations are placed in the new RMM database and flagged as less accurate and 352 locations are excluded from the database altogether (total locations RMM 8491). For IJVD 53 locations are placed in the new database and flagged as less accurate and 42 locations are excluded from the database altogether (total locations IJVD 2459).