The U.S. Geological Survey, in cooperation with the Albuquerque Bernalillo County Water Utility Authority (ABCWUA), measures groundwater levels continuously (hourly) and discretely (semiannually and annually) at a network of wells and piezometers (hereafter called the observation network) within the Middle Rio Grande Basin in central New Mexico. Groundwater levels that are measured in this observation network provide a long-term hydrologic dataset that is heavily relied upon to make water management decisions. The desire to upgrade and perform maintenance on this observation network initiated this study, which assesses the spatial and temporal importance of measurements towards optimization of monitoring the observation network to reduce or redirect monitoring costs. This report describes the methods and results of the optimization assessment of this observation network, which included separate spatial and temporal methodologies and an evaluation using principal component analysis (PCA).

Results from the spatial optimization assessment can be used to help identify observation network sites that do not significantly affect the generation of winter groundwater-elevation contour maps of the production zone. Results from the temporal optimization assessment and PCA can also be consulted when deciding which sites to remove from the network, especially for sites that are monitored more frequently than annually. Results from the temporal optimization assessment can be used to inform the minimum monitoring frequency at the observation network required to capture the trends shown in higher frequency monitoring. The PCA results distinguish spatially distributed characteristic water-level trends that can inform the management decisions that are made when using the spatial and temporal optimization assessment results. Reducing the temporal frequency or spatial density of monitoring is ultimately a management decision that depends on the amount of data loss or degradation that is deemed acceptable while still meeting the network objectives of the ABCWUA. This study can also serve as a starting point to a data gap analysis of local aquifer characteristics and help guide enhanced observation network design as needs arise or in advance of future water management decisions.

The results of the spatial optimization assessment indicate that as many as about 20 specified sites can be removed from the observation network with a relatively small loss in the ability to represent the kriged groundwater-elevation surfaces of the production zone that were generated by using median groundwater elevations for two periods: the 2001 time interval and 2015 time interval. This analysis also demonstrated the importance of wells at the margin of the study area and in areas where there are large hydrologic gradients. At many of the 47 hourly monitored sites analyzed in the temporal optimization assessment, temporal trends were well represented for at least one of the reduced monitoring frequencies tested, indicating that a reduced frequency may be sufficient to adequately characterize seasonal and long-term trends. PCA and k-means clustering analysis of the 15 hourly monitored sites that are screened within the production zone indicate that the sites can be categorized into four groups, or clusters, of differing groundwater-level hydrograph characteristics. Except for one cluster, all of the clusters have the potential to be well represented by fewer index monitoring sites.