The world evolves continuously across space and time. Massive volumes of data are generated through sensing, simulation, remote observation, and human activities, capturing dynamic processes in environmental, social, economic, and engineered systems. Critical insights are embedded within these large-scale spatiotemporal datasets. Spatiotemporal Data Science provides a conceptual and methodological framework for analyzing such data by integrating spatiotemporal thinking, computational infrastructure, artificial intelligence, and domain knowledge. The field advances methods for data acquisition, harmonization, modeling, visualization, and decision support, enabling applications in natural disaster response, public health, climate adaptation, infrastructure resilience, and geopolitical analysis. By leveraging emerging technologies—including generative Artificial Intelligence (AI), large-scale cloud platforms, Graphics Processing Unit (GPU) acceleration, and digital twin systems—Spatiotemporal Data Science enables scalable, interoperable, and solution-oriented research and innovation. It represents a critical frontier for scientific discovery, engineering advancement, technological innovation, education, and societal benefit. Spatiotemporal Data Science is a transdisciplinary field that studies and models dynamic phenomena across space and time by integrating spatial theory, temporal reasoning, artificial intelligence, and scalable computational infrastructure. It enables the development of adaptive, predictive, and increasingly autonomous systems for understanding and managing complex real-world processes.

Accurate air pollution monitoring is critical to understand and mitigate the impacts of air pollution on human health and ecosystems. Due to the limited number and geographical coverage of advanced, highly accurate sensors monitoring air pollutants, many low-cost and low-accuracy sensors have been deployed. Calibrating low-cost sensors is essential to fill the geographical gap in sensor coverage. We systematically examined how different machine learning (ML) models and open-source packages could help improve the accuracy of particulate matter (PM) 2.5 data collected by Purple Air sensors. Eleven ML models and five packages were examined. This systematic study found that both models and packages impacted accuracy, while the random training/testing split ratio (e.g., 80/20 vs. 70/30) had minimal impact (0.745% difference for R2). Long Short-Term Memory (LSTM) models trained in RStudio and TensorFlow excelled, with high R2 scores of 0.856 and 0.857 and low Root Mean Squared Errors (RMSEs) of 4.25 µg/m3 and 4.26 µg/m3, respectively. However, LSTM models may be too slow (1.5 h) or computation-intensive for applications with fast response requirements. Tree-boosted models including XGBoost (0.7612, 5.377 µg/m3) in RStudio and Random Forest (RF) (0.7632, 5.366 µg/m3) in TensorFlow offered good performance with shorter training times (<1 min) and may be suitable for such applications. These findings suggest that AI/ML models, particularly LSTM models, can effectively calibrate low-cost sensors to produce precise, localized air quality data. This research is among the most comprehensive studies on AI/ML for air pollutant calibration. We also discussed limitations, applicability to other sensors, and the explanations for good model performances. This research can be adapted to enhance air quality monitoring for public health risk assessments, support broader environmental health initiatives, and inform policy decisions.

The retrieval of cloud fraction in satellite hyperspectral sounder field of view (FOV) is crucial for numerical weather prediction. This study proposes an innovative cloud fraction retrieval model for the hyperspectral infrared sounder -Cross-track Infrared Sounder (CrIS). The model is trained with a deep neural network (DNN), using the CrIS radiation spectra as the predictors and Visible Infrared Imaging Radiometer Suite (VIIRS) cloud mask as the learning target. An ensemble of randomly selected CrIS and VIIRS data are collocated and used as the training dataset. An optimized 5-layer neural network is built to establish the relationship between the CrIS spectra and the cloud fraction calculated from the VIIRS cloud mask within the CrIS FOV. In order to reduce the number of input predictors to enhance the efficiency of the model, a principal component transformation is performed on the original CrIS spectra and only the top 77 principal component scores are adopted as the final predictors. In general, the cloud fraction retrieved from the proposed DNN model are consistent with truth values calculated from the VIIRS cloud mask product. Further analysis on use cases demonstrates a slightly better cloud retrieval result during the daytime than that of the nighttime, and ocean retrievals are more accurate than land retrievals. However, since the relationship between CrIS spectrum and the cloud fraction is nonlinear, the model tends to slightly overestimate the cloud fractions over low cloud coverage regions and underestimate the values over high cloud fraction areas. Even so, the proposed model can still be a useful tool for obtaining cloud fraction information from hyperspectral infrared sounders and has the potential to be used for the numerical weather prediction and climate models, as well as other cloud studies.

With the advancement of Artificial Intelligence (AI) technologies and accumulation of big Earth data, Deep Learning (DL) has become an important method to discover patterns and understand Earth science processes in the past several years. While successful in many Earth science areas, AI/DL applications are often challenging for computing devices. In recent years, Graphics Processing Unit (GPU) devices have been leveraged to speed up AI/DL applications, yet computational performance still poses a major barrier for DL-based Earth science applications. To address these computational challenges, we selected five existing sample Earth science AI applications, revised the DL-based models/algorithms, and tested the performance of multiple GPU computing platforms to support the applications. Application software packages, performance comparisons across different platforms, along with other results, are summarized. This article can help understand how various AI/ML Earth science applications can be supported by GPU computing and help researchers in the Earth science domain better adopt GPU computing (such as supermicro, GPU clusters, and cloud computing-based) for their AI/ML applications, and to optimize their science applications to better leverage the computing device.

Sea-level rise, and its impact on humanity, is one of the most devastating consequences of climate change. Understanding and communicating the processes that contribute to sea-level change requires observations and process models that span both global and local scales, as well as research teams that cut across the traditional Earth science disciplines. To help accelerate sea-level science and to provide data and code in an Open-Source Science framework, we are developing the Earth Information System (EIS). The EIS is a cloud-based scientific collaboration platform that provides a common computing environment, bringing process-based numerical models, NASA satellite observations, and analysis workflows together into one space. Here, we will focus on the Sea-Level Change component of the EIS; other components being prototyped currently include Fires, Freshwater, and Greenhouse Gases. The EIS Sea-Level Change component utilizes the Science Managed Cloud Environment (SMCE), a NASA-managed collection of Amazon Web Services (AWS) capabilities, as well as the Multi-Mission Algorithm and Analysis Platform (MAAP). The primary interface for researchers is a cloud-based JupyterHub, a multiuser Jupyter notebook server platform, which allows teams of researchers and software developers to co-develop and easily share model configurations and analysis workflows. For the Sea-Level Change component on EIS, we implemented a cloud-based high-performance computing (HPC) cluster and three computational models running on the cluster: the Community Firn Model (CFM), the Ice Sheet System Model (ISSM), and the Estimating the Circulation and Climate of the Ocean (ECCO) reanalysis code. Each tool can be configured and launched using a documented workflow in the interactive Jupyter notebook environment. By configuring models and analysis code to operate on the same system and by deploying workflows, our framework enables scientific collaboration by giving all users on the research team the ability to run the models and analyses, with direct access to NASA datasets in the AWS cloud. The Sea-Level Change component of the EIS has been developed in coordination with NASA's Sea-Level Change Team (N-SLCT). The N-SLCT is a team of about 80 researchers across NASA and academia working towards improving our understanding of sea-level processes and developing better sea-level projections. The N-SLCT works closely with a Practitioner Consultation Board, which is made up of members from boundary organizations, who provide guidance to the N-SLCT on how to provide useful data products and tools that can be used by boundary organizations to inform decisions. We provide several paths for researchers and users external to our team to both reproduce and extend the analyses and tools developed by our team. First, all code and notebooks developed by our research team will be made open source and shared via public git repositories. Second, some of our tools will have publicly exposed application programming interfaces (APIs), allowing others to build tools and workflows using our APIs as building blocks. By exposing certain data processing algorithms as public APIs, we hope to reduce code duplication among the entire science community. Third, we will make interactive Jupyter notebooks publicly available through the Binder Project, allowing users to spin up an interactive data analytics environment and run our notebooks. The goal of the EIS is to propel NASA Earth Science into the era of Open-Source Science, allowing all Earth Science researchers to easily access, reproduce, and extend science analyses.

Accurate weather prediction, particularly accurate temperature prediction, is critical in decision-making for energy consumption, health risks, and economics. Regional numerical weather prediction (NWP) models produce operational-level temperature forecasts based on local atmospheric circulation conditions. They suffer from dataand computational intensity, resulting in low availability of high-spatiotemporal resolution. We embed a data fusion technique into the proposed framework to address this shortcoming, integrating measurements from the Internet of Things (IoT) with a high spatiotemporal resolution with weather stations observations. The framework utilizes a Long Short-Term Memory (LSTM) network to predict surface temperature from the fusion dataset for four major cities in the U.S. (Los Angeles, New York City, Atlanta, and Chicago). The predictive framework achieves an average RMSE of 1.72 degrees C and an average R2 of 0.97 using the past 24 h to predict the future 12 h for Los Angeles (RMSE of 2.13 degrees C, 2.05 degrees C, 2.59 degrees C, and R2 of 0.98, 0.98, 0.97 for New York City, Atlanta, and Chicago). The transfer learning is adopted to leverage the pre-trained model from regions with a higher number of observation stations to predict regions with fewer stations. The transferable model improved the predicting MAE for regions with data scarcity up to 26%.

This paper investigates spatiotemporal changes of nitrogen dioxide (NO2) tropospheric vertical column density due to the COVID-19 pandemic using satellite observations before, during and after the lockdown (hereafter referred as the pre-, peri- and post-periods) in six different countries: China, South Africa, Brazil, India, the UK and the US, and compare these periods with 2019 as well as mean climatology from 2010 to 2019. We observe significant declines in relative differences (RDs) from the pre- to peri-period (as compared with the 10-year climatology) in most study countries including China, South Africa, India, and the UK by 15, 17, 8 and 7% respectively. The US does not demonstrate significant decline with RD difference relatively small at just 2%. Meanwhile, although the 2020 RD of Brazil is 7% lower than 2010-2019, this trend is quite similar to that of 2019 (20% vs 23%). In the post-period of 2020, the NO2 columns rebound in most target countries: China, US, South Africa, Brazil and UK, with similar RDs relative to the corresponding pre-period as compared with 2019 and 2010-2019. In contrast, NO in India continues to be influenced by the ongoing COVID-19 crisis with pre-to-post RD 8% lower than the average of previous 10 years. (C) 2021 Elsevier B.V. All rights reserved.

Various recent studies have shown that societal efforts to mitigate (e.g. "lockdown") the outbreak of the 2019 coronavirus disease (COVID-19) caused non-negligible impacts on the environment, especially air quality. To examine if interventional policies due to COVID-19 have had a similar impact in the US state of California, this paper investigates the spatiotemporal patterns and changes in air pollution before, during and after the lockdown of the state, comparing the air quality measurements in 2020 with historical averages from 2015 to 2019. Through time series analysis, a sudden drop and uptick of air pollution are found around the dates when shutdown and reopening were ordered, respectively. The spatial patterns of nitrogen dioxide (NO2) tropospheric vertical column density (TVCD) show a decreasing trend over the locations of major powerplants and an increasing trend over residential areas near interactions of national highways. Ground-based observations around California show a 38%, 49%, and 31% drop in the concentration of NO2, carbon monoxide (CO) and particulate matter 2.5 (PM2.5) during the lockdown (March 19-May 7) compared to before (January 26-March 18) in 2020. These are 16%, 25% and 19% sharper than the means of the previous five years in the same periods, respectively. Our study offers evidence of the environmental impact introduced by COVID-19, and insight into related economic influences.

Climate and weather data such as precipitation derived from Global Climate Models (GCMs) and satellite observations are essential for the global and local hydrological assessment. However, most climatic popular precipitation products (with spatial resolutions coarser than 10km) are too coarse for local impact studies and require “downscaling” to obtain higher resolutions. Traditional precipitation downscaling methods such as statistical and dynamic downscaling require an input of additional meteorological variables, and very few are applicable for downscaling hourly precipitation for higher spatial resolution. Based on dynamic dictionary learning, we propose a new downscaling method, PreciPatch, to address this challenge by producing spatially distributed higher resolution precipitation fields with only precipitation input from GCMs at hourly temporal resolution and a large geographical extent. Using aggregated Integrated Multi-satellitE Retrievals for GPM (IMERG) data, an experiment was conducted to evaluate the performance of PreciPatch, in comparison with bicubic interpolation using RainFARM—a stochastic downscaling method, and DeepSD—a Super-Resolution Convolutional Neural Network (SRCNN) based downscaling method. PreciPatch demonstrates better performance than other methods for downscaling short-duration precipitation events (used historical data from 2014 to 2017 as the training set to estimate high-resolution hourly events in 2018).

The advancements of sensing technologies, including remote sensing, in situ sensing, social sensing, and health sensing, have tremendously improved our capability to observe and record natural and social phenomena, such as natural disasters, presidential elections, and infectious diseases. The observations have provided an unprecedented opportunity to better understand and respond to the spatiotemporal dynamics of the environment, urban settings, health and disease propagation, business decisions, and crisis and crime. Spatiotemporal event detection serves as a gateway to enable a better understanding by detecting events that represent the abnormal status of relevant phenomena. This paper reviews the literature for different sensing capabilities, spatiotemporal event extraction methods, and categories of applications for the detected events. The novelty of this review is to revisit the definition and requirements of event detection and to layout the overall workflow (from sensing and event extraction methods to the operations and decision-supporting processes based on the extracted events) as an agenda for future event detection research. Guidance is presented on the current challenges to this research agenda, and future directions are discussed for conducting spatiotemporal event detection in the era of big data, advanced sensing, and artificial intelligence.

Detection of cloud contaminated field of views (FOV) from satellite hyperspectral infrared sounders is essential for numerical weather prediction. A new cloud detection model is developed for the cross-track infrared sounder (CrIS) using the artificial deep neural network (DNN) technique. The truth cloud information used is from another instrument of Visible Infrared Imaging Radiometer Suite (VIIRS) deployed on the same platform of CrIS. The training data set is built from CrIS–VIIRS collocated measurements randomly selected from different months to represent different atmospheric and surface conditions. Then, we use the VIIRS cloud mask collocated within the CrIS footprint to train the CrIS spectra for cloud detection. Specifically, the CrIS spectra were transformed into their principal components (PCs), with only the top 75 PCs used as the predictors rather than the entire CrIS channels, for the purpose of better regression and convergence during the training process and faster prediction. Results were examined globally by the considered truth derived from the VIIRS cloud mask. Generally, the spatial distribution of the proposed CrIS cloud detection result agrees with that from the VIIRS, with a high model accuracy of 93%. Further analysis indicates that the proposed CrIS cloud detection result is slightly better over daytime than nighttime with the accuracy values of 94% versus 91%. The ocean areas have a higher cloud detection accuracy than continental land with accuracy values of 95% versus 88%. In addition, sometimes the DNN model would recognize the thin cloud as clear sky, as their spectra are very similar to each other. False detected pixels are also found over snow- or ice-covered and desert areas. This is possibly due to the VIIRS cloud mask that has a relatively low accuracy over these areas.

Precipitation, especially convective precipitation, is highly associated with hydrological disasters (e.g., floods and drought) that have negative impacts on agricultural productivity, society, and the environment. To mitigate these negative impacts, it is crucial to monitor the precipitation status in real time. The new Advanced Baseline Imager (ABI) onboard the GOES-16 satellite provides such a precipitation product in higher spatiotemporal and spectral resolutions, especially during the daytime. This research proposes a deep neural network (DNN) method to classify rainy and non-rainy clouds based on the brightness temperature differences (BTDs) and reflectances (Ref) derived from ABI. Convective and stratiform rain clouds are also separated using similar spectral parameters expressing the characteristics of cloud properties. The precipitation events used for training and validation are obtained from the IMERG V05B data, covering the southeastern coast of the U.S. during the 2018 rainy season. The performance of the proposed method is compared with traditional machine learning methods, including support vector machines (SVMs) and random forest (RF). For rainy area detection, the DNN method outperformed the other methods, with a critical success index (CSI) of 0.71 and a probability of detection (POD) of 0.86. For convective precipitation delineation, the DNN models also show a better performance, with a CSI of 0.58 and POD of 0.72. This automatic cloud classification system could be deployed for extreme rainfall event detection, real-time forecasting, and decision-making support in rainfall-related disasters.

Earth observations and model simulations are generating big multidimensional array-based raster data. However, it is difficult to efficiently query these big raster data due to the inconsistency among the geospatial raster data model, distributed physical data storage model, and the data pipeline in distributed computing frameworks. To efficiently process big geospatial data, this paper proposes a three-layer hierarchical indexing strategy to optimize Apache Spark with Hadoop Distributed File System (HDFS) from the following aspects: (1) improve I/O efficiency by adopting the chunking data structure; (2) keep the workload balance and high data locality by building the global index (k-d tree); (3) enable Spark and HDFS to natively support geospatial raster data formats (e.g., HDF4, NetCDF4, GeoTiff) by building the local index (hash table); (4) index the in-memory data to further improve geospatial data queries; (5) develop a data repartition strategy to tune the query parallelism while keeping high data locality. The above strategies are implemented by developing the customized RDDs, and evaluated by comparing the performance with that of Spark SQL and SciSpark. The proposed indexing strategy can be applied to other distributed frameworks or cloud-based computing systems to natively support big geospatial data query with high efficiency.

AbstractThis paper describes the repackaging and consistent distribution of the world’s major atmospheric and oceanic reanalyses. It also presents examples of the usefulness of examining multiple reanalyses. This service will make it much easier for anybody using reanalysis to access multiple datasets using an approach similar to that of phase 5 of the Coupled Model Intercomparison Project (CMIP5). Experienced users as well as students will find the standardized formatted data convenient to use.