Mapping and monitoring crops is a key step towards sustainable intensification of agriculture and addressing global food security. A dataset like ImageNet that revolutionized computer vision applications can accelerate development of novel crop mapping techniques. Currently, the United States Department of Agriculture (USDA) annually releases the Cropland Data Layer (CDL) which contains crop labels at 30m resolution for the entire United States of America. While CDL is state of the art and is widely used for a number of agricultural applications, it has a number of limitations (e.g., pixelated errors, labels carried over from previous errors and absence of input imagery along with class labels). In this work, we create a new semantic segmentation benchmark dataset, which we call CalCROP21, for the diverse crops in the Central Valley region of California at 10m spatial resolution using a Google Earth Engine based robust image processing pipeline and a novel attention based spatio-temporal semantic segmentation algorithm STATT. STATT uses re-sampled (interpolated) CDL labels for training, but is able to generate a better prediction than CDL by leveraging spatial and temporal patterns in Sentinel2 multi-spectral image series to effectively capture phenologic differences amongst crops and uses attention to reduce the impact of clouds and other atmospheric disturbances. We also present a comprehensive evaluation to show that STATT has significantly better results when compared to the resampled CDL labels. We have released the dataset and the processing pipeline code for generating the benchmark dataset.
Several deep learning methods for phase retrieval exist, but most of them fail on realistic data without precise support information. We propose a novel method based on single-instance deep generative prior that works well on complex-valued crystal data.
The availability of massive earth observing satellite data provide huge opportunities for land use and land cover mapping. However, such mapping effort is challenging due to the existence of various land cover classes, noisy data, and the lack of proper labels. Also, each land cover class typically has its own unique temporal pattern and can be identified only during certain periods. In this article, we introduce a novel architecture that incorporates the UNet structure with Bidirectional LSTM and Attention mechanism to jointly exploit the spatial and temporal nature of satellite data and to better identify the unique temporal patterns of each land cover. We evaluate this method for mapping crops in multiple regions over the world. We compare our method with other state-of-the-art methods both quantitatively and qualitatively on two real-world datasets which involve multiple land cover classes. We also visualise the attention weights to study its effectiveness in mitigating noise and identifying discriminative time period.
Land cover mapping is essential for monitoring global environmental change and managing natural resources. Unfortunately, traditional classification models are plagued by limited training data available in existing land cover products and data heterogeneity over space and time. In this survey, we provide a structured and comprehensive overview of challenges in land cover mapping and machine learning methods used to address these problems. We also discuss the gaps and opportunities that exist for advancing research in this promising direction.
Streamflow prediction is one of the key challenges in the field of hydrology due to the complex interplay between multiple non-linear physical mechanisms behind streamflow generation. While physically-based models are rooted in rich understanding of the physical processes, a significant performance gap still remains which can be potentially addressed by leveraging the recent advances in machine learning. The goal of this work is to incorporate our understanding of physical processes and constraints in hydrology into machine learning algorithms, and thus bridge the performance gap while reducing the need for large amounts of data compared to traditional data-driven approaches. In particular, we propose an LSTM based deep learning architecture that is coupled with SWAT (Soil and Water Assessment Tool), an hydrology model that is in wide use today. The key idea of the approach is to model auxiliary intermediate processes that connect weather drivers to streamflow, rather than directly mapping runoff from weather variables which is what a deep learning architecture without physical insight will do. The efficacy of the approach is being analyzed on several small catchments located in the South Branch of the Root River Watershed in southeast Minnesota. Apart from observation data on runoff, the approach also leverages a 200-year synthetic dataset generated by SWAT to improve the performance while reducing convergence time. In the early phases of this study, simpler versions of the physics guided deep learning architectures are being used to achieve a system understanding of the coupling of physics and machine learning. As more complexity is introduced into the present implementation, the framework will be able to generalize to more sophisticated cases where spatial heterogeneity is present.
Causal inference is a powerful statistical methodology for explanatory analysis and individualized treatment effect (ITE) estimation, a prominent causal inference task that has become a fundamental research problem. ITE estimation, when performed naively, tends to produce biased estimates. To obtain unbiased estimates, counterfactual information is needed, which is not directly observable from data. Based on mature domain knowledge, reliable traditional methods to estimate ITE exist. In recent years, neural networks have been widely used in clinical studies. Specifically, recurrent neural networks (RNN) have been applied to temporal Electronic Health Records (EHR) data analysis. However, RNNs are not guaranteed to automatically discover causal knowledge, correctly estimate counterfactual information, and thus correctly estimate the ITE. This lack of correct ITE estimates can hinder the performance of the model. In this work we study whether RNNs can be guided to correctly incorporate ITE-related knowledge and whether this improves predictive performance. Specifically, we first describe a Causal-Temporal Structure for temporal EHR data; then based on this structure, we estimate sequential ITE along the timeline, using sequential Propensity Score Matching (PSM); and finally, we propose a knowledge-guided neural network methodology to incorporate estimated ITE. We demonstrate on real-world and synthetic data (where the actual ITEs are known) that the proposed methodology can significantly improve the prediction performance of RNN.
Most environmental data come from a minority of well-observed sites. An ongoing challenge in the environmental sciences is transferring knowledge from monitored sites to unobserved sites. Here, we demonstrate a novel transfer learning framework that accurately predicts temperature in unobserved lakes (targets) by borrowing models from highly observed lakes (sources). This method, Meta Transfer Learning (MTL), builds a meta-learning model to predict transfer performance from candidate source models to targets using lake attributes and candidates' past performance. We constructed source models at 145 well-observed lakes using calibrated process-based modeling (PB) and a recently developed approach called process-guided deep learning (PGDL). We applied MTL to either PB or PGDL source models (PB-MTL or PGDL-MTL, respectively) to predict temperatures in 305 target lakes treated as unobserved in the Upper Midwestern United States. We show significantly improved performance relative to the uncalibrated process-based General Lake Model, where the median RMSE for the target lakes is $2.52^{\circ}C$. PB-MTL yielded a median RMSE of $2.43^{\circ}C$; PGDL-MTL yielded $2.16^{\circ}C$; and a PGDL-MTL ensemble of nine sources per target yielded $1.88^{\circ}C$. For sparsely observed target lakes, PGDL-MTL often outperformed PGDL models trained on the target lakes themselves. Differences in maximum depth between the source and target were consistently the most important predictors. Our approach readily scales to thousands of lakes in the Midwestern United States, demonstrating that MTL with meaningful predictor variables and high-quality source models is a promising approach for many kinds of unmonitored systems and environmental variables.
This paper proposes a physics-guided machine learning approach that combines advanced machine learning models and physics-based models to improve the prediction of water flow and temperature in river networks. We first build a recurrent graph network model to capture the interactions among multiple segments in the river network. Then we present a pre-training technique which transfers knowledge from physics-based models to initialize the machine learning model and learn the physics of streamflow and thermodynamics. We also propose a new loss function that balances the performance over different river segments. We demonstrate the effectiveness of the proposed method in predicting temperature and streamflow in a subset of the Delaware River Basin. In particular, we show that the proposed method brings a 33\%/14\% improvement over the state-of-the-art physics-based model and 24\%/14\% over traditional machine learning models (e.g., Long-Short Term Memory Neural Network) in temperature/streamflow prediction using very sparse (0.1\%) observation data for training. The proposed method has also been shown to produce better performance when generalized to different seasons or river segments with different streamflow ranges.
In this manuscript, we provide a structured and comprehensive overview of techniques to integrate machine learning with physics-based modeling. First, we provide a summary of application areas for which these approaches have been applied. Then, we describe classes of methodologies used to construct physics-guided machine learning models and hybrid physics-machine learning frameworks from a machine learning standpoint. With this foundation, we then provide a systematic organization of these existing techniques and discuss ideas for future research.