Comparisons show high correlations of the two systems at tidal frequencies and show that, as expected, the systems track each other well on a monthly basis. Outlier differences found in monthly mean sea level and monthly mean diurnal range comparisons have been found to be partly due to; the lack of accuracy in the ADR/Bubbler times; periods of breakfills in the ADR using inferred data or backup bubbler data; and high degrees of variability in the tide staff/ETG readings. Data comparisons of over one-year in length are now being analyzed with no systematic trends or seasonal variations being found that are not explained by measurement uncertainties in the ADR/Bubbler systems. The magnitude of the differences found in the comparisons are approximately of the same order magnitude as the estimated uncertainties in the measurement systems, thus drawing conclusions on trends and seasonal variations from the differences found becomes difficult.
Both the protective well of the new system and the stilling well of the former system introduce biases in the data due to draw down effects of currents and waves. However, the new configuration (i.e., the new system's data sampling procedures in conjunction with the open protective well with parallel plates) is a significant improvement over the old. It reduces the noise present in the ADR discrete sampling method and reduces the bias due to high currents. Through the elimination of the potential for uncertainty and error using the manual staff/ETG readings and the monthly ADR/Bubbler adjustment of data using these readings, the new system eliminates the time varying biases present in the former system. The result is improved accuracy of measurement.