It is well known that the Ensemble Kalman Filter (EnKF) requires updating each ensemble member with perturbed observations to produce the proper analysis-error covariances.

This study investigates the use of non-perturbed Ensemble Kalman Filters in land data assimilation. Although these filters are suboptimal compared to the standard Ensemble Kalman Filters, in the sense that they underestimate the analysis errors, they produce smaller budget errors of conserved quantities due to the absence of artificial random perturbations.

While increased accuracy in a mean square sense may be preferable in many applications, less accuracy might be preferable in other applications, especially if the variables being assimilated obey certain conservation laws. In land data assimilation, for instance, the update in soil moisture often produces a water balance residual, in the sense that the input water is not equal to output water.

A total of eight data assimilation schemes were investigated; four of them being non-perturbed variants of previously proposed schemes. The major finding of this study is that suppressing perturbations in the EnKF significantly improves the water budget residual without significantly increasing the state errors. This finding was shown to be independent of ensemble size or observation frequency.