Browsing by Subject "Regionalization"

Now showing 1 - 2 of 2
  • Thumbnail Image
    Item
    Enabling Neighborhood Health Research and Protecting Patient Privacy
    (2021-08) Krzyzanowski, Brittany
    Maps and spatial analysis offer a more comprehensive understanding of complex neighborhood health relationships, and yet there is a remarkable lack of maps within the literature on neighborhood health. Review of the literature confirmed that only a small proportion of articles on neighborhood health (28%) contained maps. Despite this, our subsequent survey showed that the majority (63%) of investigators created maps, worked with maps, or used mapping software to explore their data at some point during their study. Neighborhood health investigators are not neglecting to explore the spatial nature of their data, but rather, they are just not publishing the maps that they are using. One of the major barriers identified by our survey was privacy regulations, such as HIPAA law, which stood as a direct barrier for 14% of survey respondents who created maps but did not share them. Many researchers find core elements of the HIPAA privacy provision specific to map data ambiguous or difficult to understand, which is reflected in disagreement and uncertainty in research and policy circles on how to enact this provision. This dissertation provides a thorough examination of the safe harbor provision and elucidates the ambiguity within the law to encourage safe and effective sharing of mapped patient data. Moreover, many scholars and policy makers have challenged this rule, saying that it is possible to share finer-grained mapped health data without jeopardizing patient privacy. One promising strategy is regionalization, or zone design, which offers a way to build finer-grained geographical units in ways that integrate the HIPAA safe harbor requirements. This dissertation explores two existing regionalization methods (Max P Regions and REDCAP) and also introduces two novel variants of these approaches (MSOM and RSOM) which we compare and contrast in terms of fitness for analysis and display of protected health information. Each regionalization procedure has its own strengths and weaknesses, but REDCAP provides the best overall performance. In general, all of the regionalization procedures produced contiguous regions that result in a better resolution map than the current standard for sharing patient data and offer to help investigators work within the bounds of privacy provisions to share maps and spatial data.
  • Thumbnail Image
    Item
    Estimating renewable water flux using landscape features.
    (2011-07) Peterson, Heidi Marie
    The complexity of vadose zone, groundwater and surface water interactions presents hydrological research challenges specifically in the area of quantifying groundwater recharge. To acknowledge unity of the surface and groundwater systems requires an interdisciplinary approach that organizes knowledge about an analysis domain based on hydrologic units rather than on aquifers, enabling an integrative, system viewpoint of the terrestrial hydrologic system. By establishing the relationship between landscape components and water balance characteristics, hydrologic response units are established. This dissertation hypothesizes that regionalization can identify hierarchical hydrogeological units (HHUs) composed of unique combinations of surface water, groundwater and vadose zone landscape characteristics with statistically different recharge rates (p<0.05), and that these units can be used to estimate the renewable water flux of the groundwater system. Three interdependent studies were pursued to address the hypothesis: (i)a statewide regionalization of mean annual streamflow that defined independent hydrologic regimes within Minnesota; (ii) a regionalization focusing on East Central Minnesota which established unique HHUs based on unique combinations of landscape characteristics with statistically significant differences in mean minimum recharge; (iii) a water management application which used the regionalization methodology to quantify the renewable groundwater flux and assess a water resources sustainability indicator within the Twin Cities Metropolitan Area. Results from the three studies indicate that landscape characteristics control the rate of renewable water flux within the groundwater system. The results identified at one analysis scale could be used to extrapolate data at refined scales where long-term hydrologic monitoring data is lacking but informed water management decisions are crucial for the sustainable future of freshwater resources. Within the state of Minnesota, five hydrologic regimes were identified, each with varying inter- and intra-annual flow characteristics. Kendall-tau results suggest that mean annual streamflow within the regimes is either increasing or remaining stable. Focusing on a smaller, regional analysis territory within East-Central Minnesota, regionalization using the Watershed Characteristics Approach (WCA) identified HHUs composed of unique combinations of hydrogeologic characteristics, bedrock material, Quaternary thickness, topography and available water capacity. These HHUs and their corresponding minimum monthly recharge rates represent the renewable groundwater flux. HHUs previously identified within the Twin Cities Metropolitan Area (TCMA) were used together with water use estimates to calculate the sustainability indicator for each community within the TCMA. Although the WCA has few assumptions, limitations of the methodology include that the hierarchical refinement is based on the availability of streamflow data, and the accuracy of the estimated recharge rates associated with each HHU is pre-determined by the resolution of spatial landscape attribute data. Future research should further evaluate the defined HHUs to confirm that they are consistent for all five of the hydrologic regimes identified within Minnesota.