Two fundamental challenges in locative Augmented Reality (AR) storytelling are collecting data about places and the lived experiences within them, and using this data to produce compelling narratives that are experienced within their original context or in new contexts. This paper introduces a multi-phased methodology that helps creators directly address both challenges. Our primary contribution is a detailed process for (1) collecting lived socio-spatial experiences and analyzing spatial configurations from a source community (e.g., a high school) using diverse, complementary methods; and (2) collaboratively synthesizing this data, with experts in situated theatre and architecture, to produce immersive AR narratives that are highly reflective of their source environment. We further show how collecting both experiential and spatial data facilitates deploying these source-based narratives in different target environments (e.g., another school or a university building). A study with 48 participants indicated that universal themes from source-synthesized stories can resonate in different contexts when the stories are deployed using our method. However, challenges remain when transferring narratives to different contexts, including contextual dissonance and complexities when applying spatial analysis metrics for adaptive deployment. This research provides a replicable methodology for the design and deployment of locative AR narratives rooted in real-world settings.