Daily surface temperatures for 185,549 lakes in the conterminous United States estimated using deep learning (1980–2020)

Willard, Jared; Read, Jordan S.; Topp, S.; Hansen, Gretchen J. A.; Kumar, Vipin

doi:10.1002/lol2.10249

Cited by 16 publications

(24 citation statements)

References 78 publications

(110 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Similarly, other machine learning tools have been developed to perform accurate air to water conversions for smaller lakes, but to date either require a training set of water in situ LSWT (Heddam et al 2020) or have used additional training variables that can be difficult to extract from point occurrence records (e.g. lake surface areas) (Willard et al 2022).…”

Section: Discussionmentioning

confidence: 99%

“…Without these data, aquatic ecologists are often left to accept the limitations of air temperature proxies and move forward with their modeling. Likewise, machine learning approaches have been developed to predict daily and summer LSWT with remarkable accuracy (Willard et al 2022), though these models are currently limited in spatial scope. However, new global‐scale LSWT databases have recently been developed that permit a large‐scale quantification of air–water temperature relationships through which one can assess the consequences of air‐for‐water temperature substitutions and develop statistical models to account for the potential mismatch.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Global maps of lake surface water temperatures reveal pitfalls of air‐for‐water substitutions in ecological prediction

Armitage

2022

Ecography

View full text Add to dashboard Cite

In modeling species distributions and population dynamics, spatially-interpolated climatic data are often used as proxies for real, on-the-ground measurements. For shallow freshwater systems, this practice may be problematic as interpolations used for surface waters are generated from terrestrial sensor networks measuring air temperatures. Using these may therefore bias statistical estimates of species' environmental tolerances or population projections -particularly among pleustonic and epilimnetic organisms. Using a global database of millions of daily satellite-derived lake surface water temperatures (LSWT), I trained machine learning models to correct for the correspondence between air and LSWT as a function of atmospheric and topographic predictors, resulting in the creation of monthly high-resolution global maps of air-LSWT offsets, corresponding uncertainty measures and derived LSWT-based bioclimatic layers for use by the scientific community. I then compared the performance of these LSWT layers and air temperature-based layers in population dynamic and ecological niche models (ENM). While generally high, the correspondence between air temperature and LSWT was quite variable and often nonlinear depending on the spatial context. These LSWT predictions were better able to capture the modeled population dynamics and geographic distributions of two common aquatic plant species. Further, ENM models trained with LSWT predictors more accurately captured lab-measured thermal response curves. I conclude that these predicted LSWT temperatures perform better than raw air temperatures when used for population projections and environmental niche modeling, and should be used by practitioners to derive more biologically-meaningful results. These global LSWT predictions and corresponding error estimates and bioclimatic layers have been made freely available to all researchers in a permanent archive.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Global maps of lake surface water temperatures reveal pitfalls of air‐for‐water substitutions in ecological prediction

Armitage

2022

Ecography

View full text Add to dashboard Cite

show abstract

“…The relatively coarse scale of ERA5-Land lake temperature estimates may obscure important differences due to, for example, lake size and topography (Deblauwe et al, 2016). If finer scale global water temperature estimates similar to those recently developed for parts of the United States (Willard et al, 2022) become available, these may substantially improve predictions. Additionally, ERA5 lake temperature estimates may be biased for regions of the globe with few long-term water temperature datasets, given the global scale of the estimates and the model focus on atmospheric climate rather than water temperature.…”

Section: Discussionmentioning

confidence: 99%

“…Our BioLake climate layers provide a global dataset for species distribution modeling of aquatic species that could be used in comparison to air temperature in future studies. Future modeling efforts could examine the use of finer scale lake and river temperature data in areas where they are available (e.g., contiguous United States; Willard et al, 2022) as a robust comparison of air versus lake temperature for AIS risk assessments.…”

Section: Discussionmentioning

confidence: 99%

BioLake: A first assessment of lake temperature‐derived bioclimatic predictors for aquatic invasive species

et al. 2023

View full text Add to dashboard Cite

Aquatic invasive species (AIS) present major ecological and economic challenges globally, endangering ecosystems and human livelihoods.Managers and policy makers thus need tools to predict invasion risk and prioritize species and areas of concern, and they often use native range climate matching to determine whether a species could persist in a new location.However, climate matching for AIS often relies on air temperature rather than water temperature due to a lack of global water temperature data layers, and predictive power of models is seldom evaluated. We developed 12 global lake (water) temperature-derived "BioLake" bioclimatic layers for distribution modeling of aquatic species and compared "climatch" climate matching predictions (from climatchR package) from BioLake with those based on BioClim temperature layers and with a null model. We did this for 73 established AIS in the United States, training the models on their ranges outside of the United States and Canada. Models using either set of climate layers outperformed the null expectation by a similar (but modest) amount on average, but some species were occasionally found in locations with low climatch scores.Mean US climatch scores were higher for most species when using air temperature. Including additional climate layers in models reduced mean climatch scores, indicating that commonly used climatch score thresholds are not absolute but can be context specific and may require calibration based upon climate data used. Although finer resolution global lake temperature data would likely improve predictions, our BioLake layers provide a starting point for aquatic species distribution modeling. Climate matching was most effective for some species that originated at low latitudes or had small ranges. Climatch scores remain useful but limited for predicting AIS risk, perhaps because current ranges seldom fully reflect climatic tolerances (fundamental niches). Managers could consider climate matching as one of a suite of tools that can be used in AIS prioritization.

show abstract

“…manuscript submitted to Water Resources Research Daily surface water temperature and corresponding weather data (wind speed) were also included in our model development. We extracted daily water temperature from Willard et al (2022), which includes estimated daily surface water temperature for 185,549 lakes across the US. In addition to daily surface temperature, we calculated prior 14-day mean temperatures for all 1,340 observations included in our training set.…”

Section: Data Sources and Processingmentioning

confidence: 99%

Long-term trends in productivity across Intermountain West lakes provide no evidence of widespread eutrophication

Sillen

Ross

Collins

2023

Preprint

View full text Add to dashboard Cite

Eutrophication represents a major threat to freshwater systems and climate change is expected to drive further increases in freshwater primary productivity. However, long-term in-situ data is available for very few lakes and makes identifying trends and drivers of eutrophication challenging. Using remote sensing data, we conducted a retrospective analysis of long-term trends in trophic status across the Intermountain West, a region with understudied water quality trends and limited long-term datasets. We found that most lakes (55%) were not exhibiting shifts in trophic status from 1984-2019. Our results also show that increases in eutrophication were rare (3% of lakes) during this period, and that lakes exhibiting negative trends in trophic status were more common (17% of lakes). Lakes that were not trending occupied a wide range of lake and landscape characteristics, whereas lakes that were becoming less eutrophic tended to be in more heavily developed catchments. Our results highlight that while there are well-established narratives that climate change can lead to more eutrophication of lakes, this is not broadly observed in our dataset, with more lakes becoming more oligotrophic than lakes becoming eutrophic.

show abstract

Daily surface temperatures for 185,549 lakes in the conterminous United States estimated using deep learning (1980–2020)

Cited by 16 publications

References 78 publications

Global maps of lake surface water temperatures reveal pitfalls of air‐for‐water substitutions in ecological prediction

Global maps of lake surface water temperatures reveal pitfalls of air‐for‐water substitutions in ecological prediction

BioLake: A first assessment of lake temperature‐derived bioclimatic predictors for aquatic invasive species

Long-term trends in productivity across Intermountain West lakes provide no evidence of widespread eutrophication

Contact Info

Product

Resources

About