The digital era has ushered in a new collective action landscape that scholars are now trying to

understand. To date, there has been insufficient attention paid to cross-national variation collective action that is driven by digitally mediated thin tie networks rather than formal political organizations. In this dissertation, I propose a theory of cross-national variation in digitally networked collective action (DNCA) and test it empirically through a paired case comparison of Japan and South Korea. I argue that the extent and scale of DNCA is to a large extent a function of citizenship norms, such that the use of digital media and related technologies will have a more transformative effect on collective action in contexts with high aggregate levels of engaged norms, and a less transformative effect in countries with low aggregate levels of engaged norms. I test this theory by looking for differences in the extent to which digitally mediated thin ties, social network size, and social network heterogeneity are conducive to participation in protests, boycotts, and petitions in Japan and South Korea. I evaluate the external validity of the results of the main analyses through a larger-N analysis.

As the global human population grows, the water demands of agriculture will likewise increase. Currently, over half of the world's water consumption is due to crop irrigation. Therefore, improving irrigation is essential to alleviating the effects of these rising demands. Developing low-cost technology for custom water delivery to individual or small groups of vines is a critical next step to advance precision irrigation. Current systems for estimating evapotranspiration (ET), or plant water use, work on the scale of a full vineyard (e.g., 3-5 acres) or the scale of a single vine, but at a cost that prohibits monitoring past a small number of representative vines. An ideal irrigation system, on the other hand, would rely on measurements of water demand—defined by both water use and water status indicators—and provide water to plants in response to these measurements.Meeting this challenge would require a multidisciplinary effort over multiple years, but that is the goal of this dissertation, to develop a comprehensive single plant water use sensing and delivery system. First, three new ET models based on first principles or simple correlations are introduced and the results of a multi-year trial at RMI vineyard are presented to illustrate the performance of each of the novel High Resolution Irrigation (HRI) models. Results suggest the three models perform well, with single vine ET measurements from all three models consistently showing a strong relationship with ground truth ET measurements. Though these early results showed HRI is possible (range in r2 = 0.07 - 0.92), in order to make this technology generalizable to any vine, it must be possible to measure the area term in each model directly from observations. This is the focus of the second chapter, in which vine physical characteristics were measured on a weekly basis throughout the season, then compared to area terms experimentally calculated periodically throughout the season using ground truth data and HRI model estimates. Multiple linear regression and principal component analyses suggest a significant relationship (p-value < 0.05, r2 = 0.58 - 0.80) between two of the model area terms and vine physical parameters including canopy superficial area, canopy polygon area and fraction of absorbed photosynthetically active radiation (fPAR). With these results it is possible to test the HRI ET models in a commercial agriculture context. Then a new low-cost biometeorological sensor is introduced, tested against reference sensors, then used to test the three HRI models in a commercial vineyard alongside research grade flux tower sensors. The Cube sensors showed strong correlation with the research grade reference sensors, fulfilling one of the early project goals of developing a low-cost biometeorological sensor for HRI and testing it in a commercial context. In the final chapter the new Leaf Area Tool image analysis pipeline is introduced, supporting HRI with new tools for extracting vine parameters related to model area terms from raw data. This field ready, fast (300x to 1400x faster than other methods tested) and accurate (r2 = 0.94 - 0.99) method for the quantification of leaf area from digital images is tested on a diverse dataset of broad leaves.

Premise

Methods and results

Conclusions