Summary

One of CAP LTER’s core research aims is to predict and explain the outcomes of social-ecological processes through understanding of past and future urban environmental change, across space and time. Documenting the spatial heterogeneity of changes in landscape and bioclimatic characteristics is fundamental to conducting such analyses. Toward this end, CAP LTER researchers have long strived to produce diverse remotely sensed geospatial datasets that complement our long-term monitoring efforts, particularly through the generation of time series products of historical and contemporary land use/land cover change (LULCC) at multiple spatial resolutions (1-m, 30-m, and 250-m) across central Arizona-Phoenix. These LULLC datasets include: 1) 1-m resolution land-cover classifications for 2010 and 2015 using NAIP (National Agricultural Imagery Program) imagery (Li et al. 2014; Zhang and Li 2020); 2) 30-m resolution land-cover classifications for 1985, 1990, 1995, 2000, 2005, 2010, 2015 and 2020 based on Landsat imagery (Zhang and Li 2017; Sabu et al. 2023); 3) historical land-use classifications of central Arizona-Phoenix 1912 to 1995 (Moritz et al. 2016); and 4) 1-m resolution “open” or “vacant” land cover for 2010 and 2015 (Smith et al. 2017; Smith et al. 2020). Additionally, researchers have created a dataset identifying key features of central Arizona’s urban ecological infrastructure (UEI), including both built and natural features of the urban landscape fundamental to the provisioning of urban ecosystem services (Brown et al. 2021).

With the proliferation in publicly available satellite imagery, particularly through platforms such as Google Earth Engine, CAP researchers have produced additional geospatial datasets for characterizing long-term and seasonal dynamics of vegetation, impervious surface cover, land surface temperature, water, and climate. This includes multiple long-term datasets of vegetation greenness (as proxy measures of vegetation productivity): 1) 30-m resolution seasonal and annual estimates of vegetation greenness indices (NDVI) for 1998 to 2023, based on Landsat imagery (Haight et al. 2025a); 2) 250-m resolution annual estimates of vegetation (EVI) seasonality from 2001 to 2018, derived from MODIS imagery (Albuquerque 2020); and 3) 1-m resolution vegetation greenness estimates (NDVI, SAVI) for 2010, 2013, 2015, 2017, 2019, and 2021, based on NAIP imagery (Stulhmacher 2019a-h; Sabu and Frazier 2023a-d). To estimate variation in impervious surface area associated with urbanization, including the creation of buildings and roads, we have estimated a variety of impervious surface indices, including the Enhanced Normalized Difference Impervious Surface Index (1998 to 2023; Haight et al 2025b). Datasets of land surface temperature (LST) include 30-m resolution time series at 5-year intervals (1985 to 2020; Frazier et al. 2021) and 1-year intervals for multiple seasons (1998 to 2023; Haight et al. 2025c). To characterize the seasonally ephemeral distributions of arid streams and waterbodies, water indices have been produced at a 30-m resolution (Haight et al. 2024c). Lastly, datasets summarizing bioclimatic conditions (air temperature, precipitation, etc.) at a 1-km resolution have been produced across central Arizona, based on NASA Daymet data (Boehme et al. 2019; Haight et al. 2025c).