Browsing by Author "Hansen, Gretchen J. A."

Now showing 1 - 2 of 2
  • Thumbnail Image
    Fish and Phytoplankton Exhibit Contrasting Temporal Species Abundance Patterns in a Dynamic North Temperate Lake
    Hansen, Gretchen J. A.; Carey, Cayelan C. (PLoS ONE, 2015-02-04)
    Temporal patterns of species abundance, although less well-studied than spatial patterns, provide valuable insight to the processes governing community assembly. We compared temporal abundance distributions of two communities, phytoplankton and fish, in a north temperate lake. We used both 17 years of observed relative abundance data as well as resampled data from Monte Carlo simulations to account for the possible effects of non-detection of rare species. Similar to what has been found in other communities, phytoplankton and fish species that appeared more frequently were generally more abundant than rare species. However, neither community exhibited two distinct groups of “core” (common occurrence and high abundance) and “occasional” (rare occurrence and low abundance) species. Both observed and resampled data show that the phytoplankton community was dominated by occasional species appearing in only one year that exhibited large variation in their abundances, while the fish community was dominated by core species occurring in all 17 years at high abundances. We hypothesize that the life-history traits that enable phytoplankton to persist in highly dynamic environments may result in communities dominated by occasional species capable of reaching high abundances when conditions allow. Conversely, longer turnover times and broad environmental tolerances of fish may result in communities dominated by core species structured primarily by competitive interactions.
  • Thumbnail Image
    Process-Guided Deep Learning Predictions of Lake Water Temperature
    Read, Jordan S.; Jia, Xiaowei; Willard, Jared; Appling, Alison P.; Zwart, Jacob A.; Oliver, Samantha K.; Karpatne, Anuj; Hansen, Gretchen J. A.; Hanson, Paul C.; Watkins, William; Steinbach, Michael; Kumar, Vipin (2019-11-08)
    The rapid growth of data in water resources has created new opportunities to accelerate knowledge discovery with the use of advanced deep learning tools. Hybrid models that integrate theory with state-of-the art empirical techniques have the potential to improve predictions while remaining true to physical laws. This paper evaluates the Process-Guided Deep Learning (PGDL) hybrid modeling framework with a use-case of predicting depth-specific lake water temperatures. The PGDL model has three primary components: a deep learning model with temporal awareness (long short-term memory recurrence), theory-based feedback (model penalties for violating conversation of energy), and model pretraining to initialize the network with synthetic data (water temperature predictions from a process-based model). In situ water temperatures were used to train the PGDL model, a deep learning (DL) model, and a process-based (PB) model. Model performance was evaluated in various conditions, including when training data were sparse and when predictions were made outside of the range in the training data set. The PGDL model performance (as measured by root-mean-square error (RMSE)) was superior to DL and PB for two detailed study lakes, but only when pretraining data included greater variability than the training period. The PGDL model also performed well when extended to 68 lakes, with a median RMSE of 1.65 degrees C during the test period (DL: 1.78 degrees C, PB: 2.03 degrees C; in a small number of lakes PB or DL models were more accurate). This case-study demonstrates that integrating scientific knowledge into deep learning tools shows promise for improving predictions of many important environmental variables.