The proliferation of machine learning models, architectures, and datasets for Earth observation (EO) continues to rise dramatically. This pattern is expected to continue growing bringing with it an increase in the generation of remote sensing derived data products powered by geospatial artificial intelligence (GeoAI) techniques. Rigorous quality assessments and accuracy analysis needs to be undertaken for the science community to adopt many of these data products for scientific discovery of changes of the Earth’s land surface. While there is existing literature supporting and documenting best practices for the validation of GeoAI data products, the software to support large-scale collaborative validation efforts is limited. In this study we present the design and software implementation of a flexible commodity-based framework for large-scale global to regional validation of GeoAI data products. This framework’s main purpose is to enable, speed up, and optimize the acquisition of validation data for large-scale science projects with support across multiple sensors and spatial resolutions with little to no code. In addition, we present several use cases where this framework has enabled and streamlined the validation of global to regional data products at different spatial resolutions and within different computational platforms.

Human-induced land cover land use (LCLU) changes such as agricultural extensification and forest degradation and loss have extensive negative impacts including biodiversity loss, land degradation, and a disruption to ecological services. In Senegal, where people are heavily reliant on dryland agricultural production, climate change and land degradation pose particularly significant threats especially as rapid population growth continues to fuel frequent LCLU change. Considering these challenges, approaches that facilitate increased insight into the spatial and temporal dynamics of land use are needed to implement sustainable land management practices and mitigation strategies. However, difficulties associated with Senegal’s highly variable phenology, sparse woody cover and small, irregular fields necessitate the use of Very High Resolution (VHR; < 3 m spatial resolution) data and modern techniques for modeling land use at sufficient scales.We take advantage of VHR data’s spatial resolution and Sentinel-1’s high temporal resolution by implementing an object-based data fusion strategy to model land use. By generating high resolution vector objects from single-date WorldView imagery and using the corresponding Synthetic Aperture Radar (SAR) time series to train a One-Dimensional Convolutional Neural Network (1D CNN), we can effectively leverage deep learning techniques to extract land use signals from multi-resolution and multi-temporal data in a near-autonomous manner.

Land cover mapping has been a valuable tool in capturing changes in many developing regions in Africa. Senegal has been a hotspot of change where agricultural activity has rapidly increased. Agriculture in this region is often a complex mosaic of small fields which makes them difficult to classify using conventional land cover mapping methods and coarse-resolution satellite imagery. WorldView (WV) satellites provide very high-resolution imagery that is ideal for semantic segmentation using convolutional neural networks (CNN). In this study, we introduced training strategies that scale up the training data for the U-Net model using 2 m WV-2 and 3 imagery to overcome the challenges of regional mapping with a patchwork of hundreds of images. The proposed strategies increased the number of training data for the U-Net model in three main scenarios, (i) conventional training, (ii) model transfer, and (iii) transfer learning, and we evaluated model generalizability on test sets for two different regions in Senegal. The results showed that models rapidly reached a high level of performance with a limited increase in additional training in conventional and transfer learning strategies. In these two strategies, the U-Net consistently produced >87% average accuracy for trained images and >70% average accuracy for all test images at the final scale level. The research opens opportunities to produce regional land cover maps in West Africa without generating a prohibitively large amount of training data.

New Observing Systems (NOS) will be NASA’s next generation approach for Earth remote sensing, utilizing many diverse observing capabilities to produce optimized measurements integrated from multiple vantage points and in multiple dimensions. NOS will require strong data fusion foundations to be able to intelligently combine, and retrieve information from, data coming from assets differing in characteristics like instrument type, spectral domain, and spatial and temporal resolution. We are developing an end-to-end data fusion framework employing advanced Artificial Intelligence (AI) Machine Learning (ML) techniques with the primary purpose to drive the design and operation of multi-sensor NOS for Earth sciences and beyond. This work requires building ML-enabled analytic tools and advanced environments to take advantage of high-performance computing systems for the creation of a NOS workflow that utilizes large amounts of diverse airborne and satellite observations along with ancillary information including climate and drought time series and soil properties. We are demonstrating the framework using a forest productivity and degradation use case, but the framework is designed to be applicable to a wide variety of NOS scientific objectives.

Scientific analysis of changes of the Earth's land surface benefit from well characterized, science quality remotely sensed data. This data quality is the result of models that estimate and remove atmospheric constituents and account for sun-sensor geometry [1] – [3]. Surface reflectance (SR) in commercial very high resolution (< 5 m; VHR) spaceborne imagery routinely varies for unchanged surface features because of signal variation from the combined effects of atmospheric haze and a range of sun-sensor geometric scenarios of acquisitions [4]. Consistency from this imagery must be sufficient to identify and track the change or stability of fine-scale features that, though small, may be widely distributed across remote domains, and serve as key indicators of critical broad-scale environmental change [5], [6]. Currently commercial SR products are available, but typically the model employed is proprietary and the costs for using these products over a large domain can be significant (e.g., Planet Surface Reflectance v.2). Here we describe an open source workflow for the scientific community to improve detection of fine-scale change with commercial VHR imagery.

Validating land cover classification results from a machine learning model is a vital step in ensuring that further decisions are based on sound and robust results that can be trusted. Calculating pixel-wise validating statistics from a stack of land cover classification results, while computationally trivial for low-resolution imagery with a small spatial footprint, poses a significant challenge for very-high-resolution (VHR) imagery spanning a larger spatial footprint. Here we describe an open-source unified Python framework and workflow for the compositing of VHR imagery based on climatic and spatial information leveraging hardware acceleration. We additionally describe the implementation of per-pixel reduction algorithms which are used to reduce the stacked composite into a robust and accurate composite that is validated.

Artificial intelligence and machine learning (AI/ML) have grown in popularity in recent decades as a result of advances in high-performance computing (HPC) and open-source software. Earth science research has benefited from these advancements and continues to do so. However, entry-level AI/ML projects frequently have a significant level of complexity that prevents them from being realized. Furthermore, HPC workflows oftentimes require specialized knowledge for the installation and management of software dependencies. Our group has been working on closing this gap, particularly focused in remotely sensed applications, by leveraging high performance computing containers to make code portable across environments. In this paper we present the design and implementation of a set of containers to both speed up and optimize the development of remotely sensed data applications. These containers have been made publicly available for the use of the science community and include common AI/ML frameworks and hardware acceleration libraries. Here we present several use cases and experiments of how containers can be leveraged to speed up the development and deployment of AI/ML software and some examples of containerized applications performance.

Using a combination of high end computing and machine learning algorithms we developed a system to interrogate climate reanalysis data in a species distribution model. The results show that this system can be used as a tool to identify key variables of interest relevant to a species and to generate a probability map of the distribution of a species of interest. This opens new avenues for statistical inference in regions with sparse observational data.

Smallholder farms in South East Asia are characterized by small irregular field patterns, dense cloud cover and haze which limits our ability to observe changes in agriculture land-use. Very-high resolution (VHR,30 m) in regions with dense persistent cloud cover and haze from biomass burning.

Commercial very high-resolution (VHR) Earth observing (EO) satellites have grown into constellations with global repeat coverage that can support existing NASA EO missions with stereo and multispectral capabilities. Sub-meter data from these instruments exceeds petabytes per year and the cost for data, storage systems and compute power have all dropped exponentially. Concurrently, through agreements with the National Geospatial-Intelligence Agency, NASA-Goddard Space Flight Center is acquiring VHR EO imagery from DigitalGlobe's WorldView-1, 2, 3 Quickbird-2, GeoEye-1 and IKONOS-2 satellites. To enhance the utility of these data we are developing an Application Program Interface (API) to produce on-demand user defined science ready products to support NASA's EO missions. These enhancements include two primary foci: 1) surface reflectance 1/2° ortho mosaics - multi-temporal 2 m multispectral imagery that can be used to investigate biodiversity, horizontal forest structure, surface water fraction, and land-cover land-use at the human scale; and 2) VHR digital elevation models (DEMs) - derived with the NASA Ames Stereo Pipeline. These enhanced products benefit Earth surface studies on the cryosphere (glacier mass balance, flow rates and snow depth), hydrology (lake/waterbody levels, landslides, subsidence) and the biosphere (vertical forest structure, tree canopy height and cover) among others. Here we present current API capabilities and recent examples of derived products used in NASA Earth Science projects.

NASA Wrangler is a loosely-coupled, event driven, highly parallel data aggregation service designed to take advantage of the elastic resource capabilities of cloud computing. Wrangler automatically collects Earth observational data, climate model outputs, derived remote sensing data products, and historic biophysical data for pre-, active-, and post-wildfire decision making. It is a core service of the RECOVER decision support system, which is providing rapid-response GIS analytic capabilities to state and local government agencies. Wrangler reduces to minutes the time needed to assemble and deliver crucial wildfire-related data.

Data from the Moderate Resolution Imaging Spectroradiometer (MODIS) represent a marked advance in global land cover mapping. The 500 m time series data have an improved spectral/spatial response when compared to heritage instruments. This study presents the first results from the Vegetation Continuous Fields algorithm of the MODIS land cover product suite. A regression tree is run using multitemporal metrics derived from a single year of MODIS data. Initial products include percent tree cover, percent short vegetation cover and percent bare ground. The results reveal the improved spatial/spectral characteristics of the MODIS data compared to the AVHRR (Advanced Very High Resolution Radiometer).

The Moderate-resolution Imaging Spectroradiometer (MODIS) instrument on board the Terra satellite offers an unprecedented combination of daily spatial coverage, spatial resolution, and spectral characteristics. These capabilities make MODIS ideal to observe a variety of rapid events: active fires, floods, smoke transport, dust storms, severe storms, iceberg calving, and volcanic eruptions. A new processing system has been developed at NASA's Goddard Space Flight Center to provide a rapid response to those events, with initial emphasis on active fire detection and 250-m resolution imagery. MODIS data of most of the Earth's land surface is processed within a few hours of data acquisition. Collaboration between NASA, the University of Maryland and the USDA Forest Service has been developed to provide fire information derived from MODIS to the fire managers. Active fire locations in the conterminous United States are produced by the MODIS Rapid Response System and communicated to the Forest Service within a few minutes of production. These active fire locations are used to generate regional fire maps, updated daily and provided to the fire managers to help them allocate adequate resources to firefighters. Active fire locations are also distributed to the Global Observation of Forest Cover (GOFC) user community through a Web interface integrating MODIS active fire locations and geographic information system datasets.