Off-campus UNL users: To download campus access dissertations, please use the following link to log into our proxy server with your NU ID and password. When you are done browsing please remember to return to this page and log out.
Non-UNL users: Please talk to your librarian about requesting this dissertation through interlibrary loan.
Drought Impacts: Detecting Deviation from Expectation across Space and Time
Scientists and policy makers need better ways to quantify effects of drought, and recognize that defining drought impacts is often subjective, relative to expectations and what is normal for a given location. This dissertation explores three questions that provide different ways of understanding drought impacts: 1) A statistical analysis to describe, explain and predict a drought signal in the annual number of human cases of West Nile Virus in Nebraska found that a dry year preceded by a wet year, often in combination with warm temperatures, increases the number of human cases. Modelling scenarios found that drought may have increased the number of cases by 38% from 2002–2018. The models were better at explaining than predicting, and better at predicting which counties would have cases (most had none) than how many cases there would be. 2) Collecting, geolocating, filtering and analyzing tweets using #drought and related hashtags, a statistical model described an expected relationship between the number of tweets in each state, each week, and news coverage, drought status, and population. I investigated whether higher-than-expected numbers of tweets, reflecting higher drought awareness, could be of use for early detection of drought, using a diversity statistic threshold to identify broader-based increases in tweet volume. Higher drought awareness in some cases reflected past drought experience, and in some cases reflected emerging drought. 3) The Drought Impact Reporter collected reports from two different networks in Missouri in 2018, one based on citizen science, and the other, tapping into stakeholders, primarily livestock producers, for crowdsourced observations. I examined differences in observations from the two networks over space, time, and drought status, and found that the citizen science reports could help validate the stakeholder reports. I conclude by evaluating how combinations of subjective, qualitative and quantitative data contribute to useful knowledge, using the Data-Information-Knowledge-Wisdom pyramid. Each of the three projects involves variables that depend on how people choose to invest time – seeking medical care, tweeting, or reporting. The final two examine whether patterns within or between groups bring meaning as we build from data to information, knowledge or wisdom, with policy implications.
Natural Resource Management|Statistics|Climate Change
Smith, Kelly Helm, "Drought Impacts: Detecting Deviation from Expectation across Space and Time" (2019). ETD collection for University of Nebraska - Lincoln. AAI27547805.