Search NASA⌕ Search

SEARCH · Search NASA

Results for “data lakes”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 613 records · Page 34

Developing reservoir monthly inflow forecasts using artificial intelligence and climate phenomenon information

Reservoirs are fundamental human–built infrastructures that collect, store, and deliver fresh surface water in a timely manner for many purposes. Efficient reservoir operation requires policy makers and operators to understand how reservoir inflows are changing under different hydrological and climatic conditions to enable forecast–informed operations. Over the last decade, the uses of Artificial Intelligence and Data Mining [AI & DM] techniques in assisting reservoir streamflow subseasonal to seasonal forecasts have been increasing. In this study, Random Forest [RF), Artificial Neural Network (ANN), and Support Vector Regression (SVR) are employed and compared with respect to their capabilities for predicting 1 month–ahead reservoir inflows for two headwater reservoirs in USA and China. Both current and lagged hydrological information and 17 known climate phenomenon indices, i.e., PDO and ENSO, etc., are selected as predictors for simulating reservoir inflows. Here, the results show three methods are capable of providing monthly reservoir inflows with satisfactory statistics; the results obtained by Random Forest have the best statistical performances compared with the other two methods; another advantage of Random Forest algorithm is its capability of interpreting raw model inputs; climate phenomenon indices are useful in assisting monthly or seasonal forecasts of reservoir inflow; and different climate conditions are autocorrelated with up to several months, and the climatic information and their lags are cross correlated with local hydrological conditions in our case studies.

54 ENVIRONMENTAL SCIENCES↗

AmeriFlux CA-GL4 Nine Mile Lighthouse

This is the AmeriFlux version of the carbon flux data for the site CA-GL4 Nine Mile Lighthouse. Site Description - Nine Mile Lighthouse is located at the south end of Simcoe Island on Lake Ontario. The eddy covariance instrumentation is located on the historic lighthouse, built in 1833, and instrumented with eddy covariance data in 2016 by a network of scientists from both US and Canada (eventually to be called the Great Lakes Evaporation Network (GLEN)). The intent of GLEN has been to provide observations of over-lake meteorology and evaporation, improve forecasting of Great Lakes water levels, and support a wide variety of stakeholders, including the National Weather Service (NWS), Environment and Climate Change Canada, National Oceanic and Atmospheric Administration, U.S. Coast Guard, recreational boaters and commercial shipping, emergency management officials, and the Great Lakes research community.

Spence, Chris [Environment and Climate Change Cana↗

Updated Subsurface Data Base For Bear Creek Valley, Chestnut Ridge, And Parts Of Bethel Valley On The U.S. Department of Energy Oak Ridge Reservation

This document represents a compilation of location, construction, and hydrologic information relating to boreholes, groundwater monitoring wells, and surface water locations that have been installed/established at the Y-12 National Security Complex (Y-12) through August 2013. To date, a total of 1422 boreholes and wells have been installed in and around the Y-12 area. Of those, 835 existing boreholes and wells continue to be utilized for groundwater monitoring programs, research, remedial investigations, plume characterization and delineation studies, and various other hydrogeologic endeavors. In addition, 215 surface water locations, such as rivers, streams, seeps, springs, lakes, ponds, and building sumps are included in this database. General data about boreholes and wells included in the database are survey coordinates, survey system, elevations, alternative names and well status. Surface water location information (Appendix I) includes name, alias, functional area, northing and easting coordinates, survey system, map number and sampling history. Tabulated construction data include total depth, completion method, borehole diameter, casing and screen materials, casing and screen diameters, casing and screen depths, filter pack depths, open-hole intervals, and open-hole diameters. Hydrogeological data summarized in this document include the aquifer monitored by the completion interval, depth to weathered and fresh bedrock, formations penetrated, well sampling history, and whether rock core and geophysical logs were obtained. This document (which is the sixth revision to Y/TS–881 and the ninth overall update of a previous compilation) is published on a regular basis by the Y-12 Groundwater Protection Program (GWPP), which serves as custodian of drilling records and well construction data for the network of wells and other groundwater monitoring stations at Y-12. The tabulations in this database are arranged in appendices of like information. An example application of the data compiled in this document would be the evaluation of the suitability of wells for continued use in groundwater investigations and monitoring studies. Other uses may include: the siting of hydrogeological characterization activities, background evaluations of sites prior to initiation of new drilling activities, and hydrogeological review of selected sites prior to the initiation of remedial actions or new construction activity. Existing boreholes, wells, and surface water locations are presented on a series of maps (Appendix K). Boreholes and wells that have been plugged and abandoned, destroyed, could not be located, or otherwise are of unknown status/condition appear on a second series of maps of identical coverage (Appendix L).

54 ENVIRONMENTAL SCIENCES↗

Cosmogenic Production of Be-7 and Be-10 in Water Targets

We have measured Be-10(t(sub 1/2) = 1.5 x 10(exp 6) years) and Be-7 (t(sub 1/2) = 53.28 days) concentrations in water targets exposed for 1 to 2 years at Echo Lake, Colorado (elevation = 3246 m) and at La Jolla, California (140 m). Neutron monitor data were used to normalize the measured concentrations in order to calculate production rates equivalent to the cosmic ray flux averaged over four solar cycles (43 years). The Be-7 production rates thus obtained correspond to 6.03 +/- 0.07 x 10(exp -6) atom/g.O/s at Echo Lake and 5.06 +/- 0.20 x 10(exp -7) atom/g.O/ s at La Jolla. The Be-10 production rates correspond to 3.14 +/- 0.18 x 10(exp -6) atom/g.O/s at Echo Lake and 2.68 +/- 0.47 x 10(exp -7) atom/g.O/s at La Jolla. When compared with Be-10 production rates determined in Be-10-saturated rocks from the Antarctic and with theoretical calculations based on meteorite and lunar sample data, we find that the million-year average production rate is about 14 - 17% greater than the present production rate averaged over the last four solar cycles. Comparison with production rates determined by measuring glacially polished rocks from the Sierra Nevada in California indicates that average production (based on a revised 13,000-year deglaciation age and a geographic latitude correction) is a about 11% greater than the average over the last four solar cycles. The measured Be-10/Be-7 production ratio in oxygen is 0.52 +/- 0.03 at Echo Lake and 0.55 +/- 0.07 at La Jolla.

Nishiizumi, K.↗

Lake Anna Water Resources: Using NASA Earth Observations to Identify Algal Event Risk Factors in Lake Anna and Help Inform Future Management Practices

Lake Anna is a man-made reservoir and popular recreation destination that spans over 13,000 acres-9,600 public and 3,400 private-in the Piedmont region of Virginia. The Lake has recently seen a rise in documented harmful algal blooms (HABs), which pose a variety of community and ecological concerns and are often exacerbated by anthropogenic factors, such as excess nutrient loads from agricultural runoff. NASA DEVELOP has partnered with the Virginia Department of Environmental Quality (DEQ) to help monitor cyanobacteria and nutrient pollution indicators across Lake Anna. The team utilized Earth observations (EO) and in situ ancillary data to identify and monitor algal bloom trends. The team used Landsat 8 Operational Land Imager (OLI), Landsat 9 OLI-2, Sentinel-2 Multispectral Instrument (MSI), and Sentinel-3 Ocean and Land Color Instrument (OLCI) to analyze water quality variables such as chlorophyll-a, turbidity, surface temperature, and cyanobacteria. Due to the lack of comprehensive in situ data and historic HAB event records, our ability to compare and validate EOs was limited. Additionally, deficient spatial resolutions, along with a geographically complicated shoreline, accentuated the spatial constraints we faced in our analysis. After examining EOs, our results indicated conditions conducive to the formation of HABs within the upper reaches of Lake Anna. Yet, spatially dependent limiting factors may have also influenced where these phenomena developed. When used in concert with existing in situ datasets, NASA EOs provide relevant stakeholders with more comprehensive analyses with which to engage in enhanced monitoring and watershed management.

cyanobacteria↗

Linking Spatiotemporal Biological Data to Predict Harmful Algal Blooms

Cyanobacterial Harmful Algal Blooms (cHABs) have significant impacts on an affected region’s economy, ecology, and human health. The blooms can release toxins that kill fish and poison water for people and animals. The global adverse effects of cHABs are exacerbated by the consequences of climate change and increased pollution. Though the phenomena are well documented, scientists’ efforts to mitigate the damage are hampered by insufficient predictive models and incomplete granular knowledge of cHAB community structure. With a goal of leveraging bioinformatics and machine learning tools to better understand and predict cHABs, we are first exploring water sample data sets. Using nearly four thousand samples from the National Center for Biotechnology Information Sequence Read Archive (NCBI-SRA) across 16 years with latitude and longitude embedded in the metadata, we mapped the location of the samples onto a Lake Erie shape file. We combined information about location, date, and community taxa in the NCBI samples to discover factors that determine cHAB features. The data are separated into three distinct zones, with the majority pooled at the southwest end of the lake and occurring in 2017. The samples are rich in biological data; our next steps are to carry out whole genome sequence analysis and use the community profiles as part of our predictive machine learning model.

59 BASIC BIOLOGICAL SCIENCES↗

Separability of boreal forest species in the Lake Jennette area, Minnesota

In order to exploit the use of thematic mapper (TM) data to obtain vegetation and net primary productivity maps in the boreal forest, three aircraft flights were undertaken over the area near Ely, MN, with the NS 001 Thematic Mapper Simulator. Attention is presently given to an analysis of these 1983 data, which attempted to separate coniferous trees from deciduous ones. Canopy reflectance models and measured optical properties of the scattering elements have been used to deepen understanding of this separability, and to relate the ratio of nadir view reflectances in TM bands 4 and 3 to the overstory leaf area index. A map that is proportional to the leaf area index for deciduous species is presented.

Shen, S. S.↗

Energy budget measurements using eddy correlation and Bowen ratio techniques at the Kinosheo Lake tower site during the Northern Wetlands Study

Fluxes of heat and water vapor were measured on a 20-m tower at Kinosheo Lake in the Hudson Bay lowlands using eddy correlation and Bowen ratio energy balance techniques. The study period was June 25 to July 28, 1990. Measurements were made over a peat bog consisting of a mixture of sphagnum moss and lichen hummocks and black pools. About 200 m west of the tower were several shallow ponds. The hummocks had a dry, insulating surface and were underlain by an ice layer near 50 cm depth until mid-July. At the beginning of the period the black pools were covered with water, and although the free water gradually disappeared over the study period, they remained saturated to the end of July. The depth of peat near the tower was about 3 m. Despite the ice layer under the hummocks, their daytime surface temperatures were high, near 35 C, and after the middle of July, above 40 C. Inspection of temperature, precipitation, and radiation data showed that the midsummer period of 1990 was warmer, drier, and sunnier than usual at Moosonee and so by influence at Lake Kinosheo. When all the data were combined to yield average diurnal energy balance components, the eddy correlation fluxes accounted for 90% of the available energy. Latent heat flux averaged 46% of the total available energy and the sensible heat flux averaged 34%. Daytime Bowen ratios were near 1 for the experimental period, suggesting that the bog behaved more like a dryland than a wetland. Eddy correlation measurements of sensible heat and latent heat flux were less than those measured using the Bowen ratio energy balance technique, the average ratios being 0.81 and 0.86 respectively. These differences were possibly due to the difficulty in measuring energy balance components of net radiation and ground heat flux over the mosaic surface.

Den Hartog, G.↗

Whole‐lake invasive crayfish removal and qualitative modeling reveal habitat‐specific food web topology

Abstract Patterning of the presence/absence of food web linkages (hereafter topology) is a fundamental characteristic of ecosystems that can influence species responses to perturbations. However, the insight from food web topology into dynamic effects of perturbations on species is potentially hindered because most described topologies represent data integrated across spatial and temporal scales. We conducted a 10‐year, whole‐lake experiment in which we removed invasive rusty crayfish ( Orconectes rusticus ) from a 64‐ha north‐temperate lake and monitored responses of multiple trophic levels. We compared species responses observed in two sub‐habitats to the responses predicted from all topologies of an integrated, literature‐informed base food web model of 32 potential links. Out of 4.3 billion possible topologies, only 308,833 (0.0072%) predicted responses that qualitatively matched observed species responses in cobble habitat, and only 12,673 (0.0003%) matched observed responses in sand habitat. Furthermore, when constrained to predictions that both matched observed responses and were highly reliable (i.e., predictions were robust to link strength values), only 5040 (0.0001%) and 140 (0.000003%) topologies were identified for cobble and sand habitats, respectively. A small number of linkages were nearly always present in these valid, reliable networks in sand, while a greater variety of possible network configurations were possible in cobble. Direct links involving invasive rusty crayfish were more important in cobble, while indirect effects involving Lepomis spp. were more important in sand. Importantly, the importance of individual species linkages differed dramatically among cobble and sand sub‐habitats within a single lake, even though species composition was identical. Although the true topology of food webs is difficult to determine, constraining topologies to include spatial resolution that matches observed experimental outcomes may reduce possibilities to a small number of plausible alternatives.

59 BASIC BIOLOGICAL SCIENCES↗

The PRISM4 (mid-Piacenzian) Palaeoenvironmental Reconstruction

The mid-Piacenzian is known as a period of relative warmth when compared to the present day. A comprehensive understanding of conditions during the Piacenzian serves as both a conceptual model and a source for boundary conditions as well as means of verification of global climate model experiments. In this paper we present the PRISM4 reconstruction, a paleoenvironmental reconstruction of the mid-Piacenzian (approximately 3 Ma) containing data for paleogeography, land and sea ice, sea-surface temperature, vegetation, soils, and lakes. Our retrodicted paleogeography takes into account glacial isostatic adjustments and changes in dynamic topography. Soils and lakes, both significant as land surface features, are introduced to the PRISM reconstruction for the first time. Sea-surface temperature and vegetation reconstructions are unchanged but now have confidence assessments. The PRISM4 reconstruction is being used as boundary condition data for the Pliocene Model Intercomparison Project Phase 2 (PlioMIP2) experiments.

soils↗

Evaluation of computer-aided procedure for detecting surface water

Results from an evaluation of a computer-aided procedure for processing ERTS-1 data to detect and locate surface water are presented. The procedure was evaluated using data from a study area in the vicinity of the Lake Somerville area in Washington County, Texas. The procedure consisted of (1) selecting water training fields, (2) aggregating the training samples together and clustering them into unimodal clusters, (3) computing the mean vector and covariance matrix for each cluster, (4) classifying all of the study area into classes corresponding to the clusters using the maximum likelihood classifier, and (5) thresholding out the nonwater pixels.

Source record↗

Part 2: Sedimentary geology of the Valles, Marineris, Mars and Antarctic dry valley lakes

Detailed mapping of the layered deposits in the Valles Marineris, Mars from high-resolution Viking orbiter images revealed that they from plateaus of rhythmically layered material whose bases are in the lowest elevations of the canyon floors, and whose tops are within a few hundred meters in elevation of the surrounding plateaus. Four hypotheses for the origin of the layered deposits were considered: that they are eolian deposits; that they are remnants of the same material as the canyon walls; that they are explosive volcanic deposits; or that they were deposited in standing bodies of water. There are serious morphologic objections to each of the first three. The deposition of the layered deposits in standing bodies of water best explains their lateral continuity, horizontality, great thickness, rhythmic nature, and stratigraphic relationships with other units within the canyons. The Martian climatic history indicated that any ancient lakes were ice covered. Two methods for transporting sediment through a cover of ice on a martian lake appear to be feasible. Based on the presently available data, along with the theoretical calculations presented, it appears most likely that the layered deposits in the Valles Marineris were laid down in standing bodies of water.

Nedell, Susan S.↗

Paleomagnetic and structural evidence for localized tectonic rotation associated with fault drag in the northeastern Mojave Desert: Implications for the late Cenozoic tectonic evolution of the Eastern California shear zone

Paleomagnetic data, coupled with detailed geological analysis of the southeastern Goldstone Lake region, indicate that lower Miocene volcanic and epiclastic rocks of the Pink Canyon area have been folded, faulted and tectonically rotated approximately 28.4 + or - 9.0 deg clockwise about a vertical axis. The Coyote Canyon Fault is locally folded about a vertical axis approximately 25 deg in a clockwise sense in the Pink Canyon area. Timing relationships indicate that rotation is post-early Miocene; regional relationships imply that deformation is late Miocene to Holocene in age. These relationships imply that tectonic rotation is local rather than regional in extent as proposed by some tectonic models. The results of this study are generally consistent with the Dokka and Travis model and the subsequently revised Dokka model of strain partitioning in the northeastern Mojave Desert block.

Macconnell, D. F.↗

The Transition of High-Resolution NASA MODIS Sea Surface Temperatures into the WRF Environmental Modeling System

The NASA Short-term Prediction Research and Transition (SPoRT) Center has developed a Moderate Resolution Imaging Spectroradiometer (MODIS) sea surface temperature (SST) composite at 2-km resolution that has been implemented in version 3 of the National Weather Service (NWS) Weather Research and Forecasting (WRF) Environmental Modeling System (EMS). The WRF EMS is a complete, full physics numerical weather prediction package that incorporates dynamical cores from both the Advanced Research WRF (ARW) and the Non-hydrostatic Mesoscale Model (NMM). The installation, configuration, and execution of either the ARW or NMM models is greatly simplified by the WRF EMS to encourage its use by NWS Weather Forecast Offices (WFOs) and the university community. The WRF EMS is easy to run on most Linux workstations and clusters without the need for compilers. Version 3 of the WRF EMS contains the most recent public release of the WRF-NMM and ARW modeling system (version 3 of the ARW is described in Skamarock et al. 2008), the WRF Pre-processing System (WPS) utilities, and the WRF Post-Processing program. The system is developed and maintained by the NWS National Science Operations Officer Science and Training Resource Coordinator. To initialize the WRF EMS with high-resolution MODIS SSTs, SPoRT developed the composite product consisting of MODIS SSTs over oceans and large lakes with the NCEP Real-Time Global (RTG) filling data over land points. Filling the land points is required due to minor inconsistencies between the WRF land-sea mask and that used to generate the MODIS SST composites. This methodology ensures a continuous field that adequately initializes all appropriate arrays in WRF. MODIS composites covering the Gulf of Mexico, western Atlantic Ocean and the Caribbean are generated daily at 0400, 0700, 1600, and 1900 UTC corresponding to overpass times of the NASA Aqua and Terra polar orbiting satellites. The MODIS SST product is output in gridded binary-1 (GRIB-1) data format for a seamless incorporation into WRF via the WPS utilities. The full-resolution, 1-km MODIS product is sub-sampled to 2-km grid spacing due to limitations in handling very large dimensions in the GRIB-1 data format. The GRIB-1 files are posted online at ftp://ftp.nsstc.org/sstcomp/WRF/, which is directly accessed by the WRF EMS scripts. The MODIS SST composites are also downloaded to the EMS data server, which is accessible by the WRF EMS users and NWS WFOs. The SPoRT MODIS SST composite provides the model with superior detail of the ocean gradients around Florida and surrounding waters, whereas the operational RTG SST typically depicts a relatively smooth field and is not able to capture sharp horizontal gradients in SST. Differences of 2-3 C are common over small horizontal distances, leading to enhanced SST gradients on either side of the Gulf Stream and along the edges of the cooler shelf waters. These sharper gradients can in turn produce atmospheric responses in simulated temperature and wind fields as depicted in LaCasse et al. Differences in atmospheric verification statistics over a several month study were generally small in the vicinity of south Florida; however, the validation of SSTs at specific buoy locations revealed important improvements in the biases and RMS errors, especially in the vicinity of the cooler shelf waters off the east-central Florida coast. A current weakness in the MODIS SST product is the occurrence of occasional discontinuities caused by high latency in SST coverage due to persistent cloud cover. An enhanced method developed by Jedlovec et al. (2009, GHRSST User Symposium) reduces the occurrence of these problems by adding Advanced Microwave Scanning Radiometer -- EOS (AMSR-E) SST data to the compositing process. Enhanced SST composites are produced over the ocean regions surrounding the Continental U.S. at four times each day corresponding to Terra and Aqua equator crossing times. For a given day and overpass time, both MODInd AMSR-E data from the previous seven days form a collection used in the compositing. At each MODIS pixel, cloud-free SST values from the collection are used to form a weighted average based on their latency (number of days from the current day). In this way, recent SST data are given more weight than older data. One of the primary issues involved in incorporating the AMSR-E microwave data in the composites is the tradeoff between the decreased spatial resolution of the AMSR-E data (25 km) and the increased coverage due to its near all-weather capability. Currently, the AMSR-E is given a weight of 20% compared to MODIS data, thereby preserving the spatial structure observed in the MODIS data. Day-time (night-time) AMSR-E SST data from Aqua are used with both Terra and Aqua MODIS day-time (night-time) SST data sets.

Case, Jonathan L.↗

Hyperspectral Imager for the Coastal Ocean (HICO): Overview, Operational Updates, and Coastal Ocean Applications

The Hyperspectral Imager for the Coastal Ocean (HICO) was built to measure in‐water properties of complex coastal regions. HICO enables synoptic coverage; 100‐meter spatial resolution for sampling the variability and spatial irregularity of coastal waters; and high spectral resolution to untangle the signals from chlorophyll, colored dissolved organic matter, suspended sediments and varying bottom types. HICO was built by the Naval Research Laboratory, installed on the International Space Station (ISS) in September 2009, and operated for ONR for the first three years. In 2013, NASA assumed sponsorship of operations in order to leverage HICO's ability to address their Earth monitoring mission. This has opened up access of HICO data to the broad research community. Over 8000 images are now available on NASA's Ocean Color Website (http://oceancolor.gsfc.nasa.gov/cgi/browse.pl?sen=hi). Oregon State University's HICO website (http://hico.coas.oregonstate.edu) remains the portal for researchers to request new collections and access their requested data. We will present updates on HICO's calibration and improvements in geolocation and show examples of the use of HICO data to address issues in the coastal ocean and Great Lakes.

Davis, Curtiss O.↗

Modeling and Observations of Outlet Canyons from Lake Overflow Floods on Early Mars

Numerous observations from both orbital remote sensing [1-3] and Mars Curiosity [4] suggest that lakes were once part of the martian landscape. From orbital data, one of the key lines of evidence for past paleolakes is the existence of several hundred valley network-fed basins – usually craters – that have outlet valleys that remain perched above their floors. The existence of outlets requires that water ponded to the point that it overflowed confining topography. Beyond recognizing these landforms, there has been only limited work reconstructing the morphometry, formative hydrology, and incision history for these outlets. Here, we describe our recently published observations of outlets and ongoing numerical modeling looking at these factors.

Fassett, Caleb I.↗

Desiccation cracks provide evidence of lake drying on Mars, Sutton Island member, Murray formation, Gale Crater

Mars Science Laboratory (MSL) Curiosity rover data are used to describe the morphology of desiccation cracks observed in ancient lacustrine strata at Gale crater, Mars, and to interpret their paleoenvironmental setting. The desiccation cracks indicate subaerial exposure of lacustrine facies in the Sutton Island member of the Murray formation. In association with ripple cross-stratification and possible eolian cross-bedding, these facies indicate a transition from longer-lived perennial lakes recorded by older strata to younger lakes characterized by intermittent exposure. The transition from perennial to episodically exposed lacustrine environments provides evidence for local to regional climate change that can help constrain Mars climate models.

58 GEOSCIENCES↗

Integrating Deep Learning and Hydrodynamic Modeling to Improve the Great Lakes Forecast

The Laurentian Great Lakes, one of the world’s largest surface freshwater systems, pose a modeling challenge in seasonal forecast and climate projection. While physics-based hydrodynamic modeling is a fundamental approach, improving the forecast accuracy remains critical. In recent years, machine learning (ML) has quickly emerged in geoscience applications, but its application to the Great Lakes hydrodynamic prediction is still in its early stages. This work is the first one to explore a deep learning approach to predicting spatiotemporal distributions of the lake surface temperature (LST) in the Great Lakes. Our study shows that the Long Short-Term Memory (LSTM) neural network, trained with the limited data from hypothetical monitoring networks, can provide consistent and robust performance. The LSTM prediction captured the LST spatiotemporal variabilities across the five Great Lakes well, suggesting an effective and efficient way for monitoring network design in assisting the ML-based forecast. Furthermore, we employed an explainable artificial intelligence (XAI) technique named SHapley Additive exPlanations (SHAP) to uncover how the features impact the LSTM prediction. Our XAI analysis shows air temperature is the most influential feature for predicting LST in the trained LSTM. The relatively large bias in the LSTM prediction during the spring and fall was associated with substantial heterogeneity of air temperature during the two seasons. In contrast, the physics-based hydrodynamic model performed better in spring and fall yet exhibited relatively large biases during the summer stratification period. Finally, we developed a statistical integration of the hydrodynamic modeling and deep learning results based on the Best Linear Unbiased Estimator (BLUE). The integration further enhanced prediction accuracy, suggesting its potential for next-generation Great Lakes forecast systems.

Xue, Pengfei (ORCID:000000025702421X)↗