Over the last several decades, the study of Earth surface processes has progressed from a descriptive science to an increasingly quantitative one due to advances in theoretical, experimental, and computational geosciences. The importance of geomorphic forecasts has never been greater, as technological development and global climate change threaten to reshape the landscapes that support human societies and natural ecosystems. Here we explore best practices for developing socially-relevant forecasts of Earth surface change, a goal we are calling “earthcasting”. We suggest that earthcasts have the following features: they focus on temporal (~1 to ~100 years) and spatial (~1 m to ~10 km) scales relevant to planning; they are designed with direct involvement of stakeholders and public beneficiaries through the evaluation of the socioeconomic impacts of geomorphic processes; and they generate forecasts that are clearly stated, testable, and include quantitative uncertainties. Earthcasts bridge the gap between Earth surface researchers and decision-makers, stakeholders, researchers from other disciplines, and the general public. We investigate the defining features of earthcasts and evaluate some specific examples. This paper builds on previous studies of prediction in geomorphology by recommending a roadmap for (i) generating earthcasts, especially those based on modeling; (ii) transforming a subset of geomorphic research into earthcasts; and (iii) communicating earthcasts beyond the geomorphology research community. Earthcasting exemplifies the social benefit of geomorphology research, and it calls for renewed research efforts toward further understanding the limits of predictability of Earth surface systems and processes, and the uncertainties associated with modeling geomorphic processes and their impacts.

Small Island Nations are at considerable risk of climate change impacts from sea-level rise to coral acidification to increasing cyclone intensity; understanding how they will change in the coming century is vital for climate mitigation and resiliency. Atoll morphometrics are calculated for 3,795 motu and 593 reef flats on 154 atolls. The total land (motu) area is 1,836.55 km2 with a total reef flat area of 7,387.43 km2. A consistent methodology to classify, segment, and calculate morphometrics is used. Composites are created for 4 years (2015- 2018), and are classified into motu, reef flat, open water/lagoon via unsupervised classification. Morphometrics are computed for each motu and reef flat of the atoll in python, creating a database of atolls and their associated morphometrics. Consistency in processing removes spatial and user bias, enabling a better understanding of geographic patterns of atolls. We identify trends in atoll, motu, and reef flat formations. The average atoll reef flat width is 850 ± 817 m and the average motu width is 263 ± 210 m. Distinct differences in the distribution of motu can be seen on a regional scale in French Polynesia, while globally, wider reef flats with larger motu are found closest to the equator. Globally there is a consistent reef flat width in front of large motu (> 10 km length) of 188 ±156 m. Our atoll morphometric database creates a baseline of current atoll characteristics that can be expanded upon in the future and used for evaluating temporal changes to atoll islands.