An official website of the United States government
Abstract. Deep learning (DL) rainfall–runoff models outperform conceptual, process-based models in a range of applications. However, it remains unclear whether DL models can produce physically plausible projections of streamflow under climate change. We investigate this question through a sensitivity analysis of modeled responses to increases in temperature and potential evapotranspiration (PET), with other meteorological variables left unchanged. Previous research has shown that temperature-based PET methods overestimate evaporative water loss under warming compared with energy budget-based PET methods. We therefore assume that reliable streamflow responses to warming should exhibit less evaporative water loss when forced with smaller, energy-budget-based PET compared with temperature-based PET. We conduct this assessment using three conceptual, process-based rainfall–runoff models and three DL models, trained and tested across 212 watersheds in the Great Lakes basin. The DL models include a Long Short-Term Memory network (LSTM), a mass-conserving LSTM (MC-LSTM), and a novel variant of the MC-LSTM that also respects the relationship between PET and evaporative water loss (MC-LSTM-PET). After validating models against historical streamflow and actual evapotranspiration, we force all models with scenarios of warming, historical precipitation, and both temperature-based (Hamon) and energy-budget-based (Priestley–Taylor) PET, and compare their responses in long-term mean daily flow, low flows, high flows, and seasonal streamflow timing. We also explore similar responses using a national LSTM fit to 531 watersheds across the United States to assess how the inclusion of a larger and more diverse set of basins influences signals of hydrological response under warming. The main results of this study are as follows: The three Great Lakes DL models substantially outperform all process-based models in streamflow estimation. The MC-LSTM-PET also matches the best process-based models and outperforms the MC-LSTM in estimating actual evapotranspiration. All process-based models show a downward shift in long-term mean daily flows under warming, but median shifts are considerably larger under temperature-based PET (−17 % to −25 %) than energy-budget-based PET (−6 % to −9 %). The MC-LSTM-PET model exhibits similar differences in water loss across the different PET forcings. Conversely, the LSTM exhibits unrealistically large water losses under warming using Priestley–Taylor PET (−20 %), while the MC-LSTM is relatively insensitive to the PET method. DL models exhibit smaller changes in high flows and seasonal timing of flows as compared with the process-based models, while DL estimates of low flows are within the range estimated by the process-based models. Like the Great Lakes LSTM, the national LSTM also shows unrealistically large water losses under warming (−25 %), but it is more stable when many inputs are changed under warming and better aligns with process-based model responses for seasonal timing of flows. Ultimately, the results of this sensitivity analysis suggest that physical considerations regarding model architecture and input variables may be necessary to promote the physical realism of deep-learning-based hydrological projections under climate change.
more »
« less
This content will become publicly available on February 1, 2026
Generalization Enhancement Strategies to Enable Cross-Year Cropland Mapping with Convolutional Neural Networks Trained Using Historical Samples
Mapping agricultural fields using high-resolution satellite imagery and deep learning (DL) models has advanced significantly, even in regions with small, irregularly shaped fields. However, effective DL models often require large, expensive labeled datasets, which are typically limited to specific years or regions. This restricts the ability to create annual maps needed for agricultural monitoring, as changes in farming practices and environmental conditions cause domain shifts between years and locations. To address this, we focused on improving model generalization without relying on yearly labels through a holistic approach that integrates several techniques, including an area-based loss function, Tversky-focal loss (TFL), data augmentation, and the use of regularization techniques like dropout. Photometric augmentations helped encode invariance to brightness changes but also increased the incidence of false positives. The best results were achieved by combining photometric augmentation, TFL, and Monte Carlo dropout, although dropout alone led to more false negatives. Input normalization also played a key role, with the best results obtained when normalization statistics were calculated locally (per chip) across all bands. Our U-Net-based workflow successfully generated multi-year crop maps over large areas, outperforming the base model without photometric augmentation or MC-dropout by 17 IoU points.
more »
« less
- Award ID(s):
- 2439879
- PAR ID:
- 10632078
- Publisher / Repository:
- MDPI
- Date Published:
- Journal Name:
- Remote Sensing
- Volume:
- 17
- Issue:
- 3
- ISSN:
- 2072-4292
- Page Range / eLocation ID:
- 474
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
In this paper we proposed a real-time face mask detection and recognition for CCTV surveillance camera videos. The proposed work consists of six steps: video acquisition and keyframes selection, data augmentation, facial parts segmentation, pixel-based feature extraction, Bag of Visual Words (BoVW) generation, face mask detection, and face recognition. In the first step, a set of keyframes are selected using histogram of gradient (HoG) algorithm. Secondly, data augmentation is involved with three steps as color normalization, illumination correction (CLAHE), and poses normalization (Angular Affine Transformation). In third step, facial parts are segmented using clustering approach i.e. Expectation Maximization with Gaussian Mixture Model (EM-GMM), in which facial regions are segmented into Eyes, Nose, Mouth, Chin, and Forehead. Then, Pixel-based Feature Extraction is performed using Yolo Nano approach, which performance is higher and lightweight model than the Yolo Tiny V2 and Yolo Tiny V3, and extracted features are constructed into Codebook by Hassanat Similarity with K-Nearest neighbor (H-M with KNN) algorithm. For mask detection, L2 distance function is used. The final step is face recognition which is implemented by a Kernel-based Extreme Learning Machine with Slime Mould Optimization (SMO). Experiments conducted using Python IDLE 3.8 for the proposed Yolo Nano model and also previous works as GMM with Deep learning (GMM+DL), Convolutional Neural Network (CNN) with VGGF, Yolo Tiny V2, and Yolo Tiny V3 in terms of various performance metrics.more » « less
-
Deep learning (DL) is revolutionizing many fields. However, there is a major bottleneck for the wide adoption of DL: the pain of model selection , which requires exploring a large config space of model architecture and training hyper-parameters before picking the best model. The two existing popular paradigms for exploring this config space pose a false dichotomy. AutoML-based model selection explores configs with high-throughput but uses human intuition minimally. Alternatively, interactive human-in-the-loop model selection completely relies on human intuition to explore the config space but often has very low throughput. To mitigate the above drawbacks, we propose a new paradigm for model selection that we call intermittent human-in-the-loop model selection . In this demonstration, we will showcase our approach using five real-world DL model selection workloads. A short video of our demonstration can be found here: https://youtu.be/K3THQy5McXc.more » « less
-
Nutrient runoff from agricultural regions of the midwestern U.S. corn belt has degraded water quality in many inland and coastal water bodies such as the Great Lakes and Gulf of Mexico. Under current climate, observational studies have shown that winter cover crops can reduce dissolved nitrogen and phosphorus losses from row-cropped agricultural watersheds, but performance of cover crops in response to climate variability and climate change has not been systematically evaluated. Using the Soil & Water Assessment Tool (SWAT) model, calibrated using multiple years of field-based data, we simulated historical and projected future nutrient loss from two representative agricultural watersheds in northern Indiana, USA. For 100% cover crop coverage, historical simulations showed a 31–33% reduction in nitrate (NO3−) loss and a 15–23% reduction in Soluble Reactive Phosphorus (SRP) loss in comparison with a no-cover-crop baseline. Under climate change scenarios, without cover crops, projected warmer and wetter conditions strongly increased nutrient loss, especially in the fallow period from Oct to Apr when changes in infiltration and runoff are largest. In the absence of cover crops, annual nutrient losses for the RCP8.5 2080s scenario were 26–38% higher for NO3−, and 9–46% higher for SRP. However, the effectiveness of cover crops also increased under climate change. For an ensemble of 60 climate change scenarios based on CMIP5 RCP4.5 and RCP8.5 scenarios, 19 out of 24 ensemble-mean simulations of future nutrient loss with 100% cover crops were less than or equal to historical simulations with 100% cover crops, despite systematic increases in nutrient loss due to climate alone. These results demonstrate that planting winter cover crops over row-cropped land areas constitutes a robust climate change adaptation strategy for reducing nutrient losses from agricultural lands, enhancing resilience to a projected warmer and wetter winter climate in the midwestern U.S.more » « less
-
Smart IoT-based systems often desire continuous execution of multiple latency-sensitive Deep Learning (DL) appli- cations. The edge servers serve as the cornerstone of such IoT- based systems, however, their resource limitations hamper the continuous execution of multiple (multi-tenant) DL applications. The challenge is that, DL applications function based on bulky “neural network (NN) models” that cannot be simultaneously maintained in the limited memory space of the edge. Accordingly, the main contribution of this research is to overcome the memory contention challenge, thereby, meeting the latency constraints of the DL applications without compromising their inference accuracy. We propose an efficient NN model management frame- work, called Edge-MultiAI, that ushers the NN models of the DL applications into the edge memory such that the degree of multi-tenancy and the number of warm-starts are maximized. Edge-MultiAI leverages NN model compression techniques, such as model quantization, and dynamically loads NN models for DL applications to stimulate multi-tenancy on the edge server. We also devise a model management heuristic for Edge-MultiAI, called iWS-BFE, that functions based on the Bayesian theory to predict the inference requests for multi-tenant applications, and uses it to choose the appropriate NN models for loading, hence, increasing the number of warm-start inferences. We evaluate the efficacy and robustness of Edge-MultiAI under various configurations. The results reveal that Edge-MultiAI can stimulate the degree of multi-tenancy on the edge by at least 2× and increase the number of warm-starts by ≈ 60% without any major loss on the inference accuracy of the applications.more » « less
