Climate change coupled with urbanization represents the biggest challenge of our generation, nationally and globally. To address this key Global Grand Challenge, it is urgent to better understand urban land cover land use change, its impacts on cities across the globe, and its interactions with climate systems across scales. Cities do not necessarily have to be a problem, but rather, should be a key to solutions, with the ability to shield urban residents from broader scale climate change. Realizing this goal, however, requires advanced data and tools, both to better understand urbanization and their impacts and for planning effective climate adaptation and mitigation strategies. Such tools, however, are largely lagging behind due to three critical barriers: (i) missing global high-resolution urban surface constraints for Earth system modeling, (ii) inability to simulate dynamic urban land change in state-of-the-art Earth system models (ESMs), and (iii) limited high-resolution modeling capabilities over urban landscapes in advanced ESMs and Earth System Digital Twins (ESDTs). Here we propose to leverage a suite of satellite observations and satellite-derived products to represent spatiotemporally continuous biophysical properties of urban areas in Earth system models to advance high-resolution urban modeling capabilities at large scales. The specific objectives of the proposed project are: (i) to advance spatiotemporally explicit urban representation in ESMs by providing multiple realizations of high-resolution (~1 km) urban land surface datasets as boundary conditions, and (ii) to develop km-scale climate modeling capability of urban land cover land use change (LCLUC). To achieve these objectives, we design a cohesive workflow to develop the first-of-its-kind km-scale transient urban land surface datasets leveraging remote sensing products and machine learning techniques, and to model the climatic impacts of urbanization across scales with assessment of the associated robustness and uncertainties, primarily focused on the Community Earth System Model and the Energy Exascale Earth System Model. We will focus on the inclusion of various geospatial and satellite-derived data sources to develop multiple realizations of these urban surface constraints. This will generate a range of historical urban climate simulations and provide statistically robust estimates, which is critical for complex non-linear systems.

Rapid urban development in the future will subject urban areas and their residents to substantial climate risks, but also presents a historic and time-sensitive opportunity to mitigate the negative impacts of climate change and urban growth. The proposed work is both broadly significant and timely -- occurring at the confluence of two of the most dominant forces shaping the globe today, urbanization and climate change. The proposed research is closely relevant to the NASA LCLUC Program and directly addresses the sub-element 1 Land Use for Digital Twins (LU4DT) of the solicitation, specifically the Type 2 proposal of this sub-element -- "incorporation of land-use datasets as boundary conditions as a function of time, e.g. on an annual basis, in multiannual climate model runs for the last decade or longer and comparison of the hindcast results with the observed climate variables". Outcomes of this work will be essential to accurately resolve urban climate impacts in ESMs. Datasets created from this project will all be publicly released and promote high-resolution (km-scale) Earth system modeling. Additionally, with the incorporation of more human systems within ESMs through representation of urban areas, this will lead to the development of complete Land-Earth System Digital Twins. The results and insights derived from the proposed work would contribute to advance the fundamental understanding of how large-scale climate variability and change coupled with urbanization affects the urban environments across scales.