Large-scale Meteorological Data Research Articles

Bringing vision to climate: A hierarchical model for water depth monitoring in headwater streams

Stream-observing cameras have recently been deployed in stream systems to monitor water depth dynamics. However, most existing image-based water depth monitoring methods require additional gauging equipment, extensive manual annotations, or complex manual calibration. In this paper, we propose the hierarchical model, a novel multi-modal and multi-scale deep learning framework for monitoring water depth in headwater streams with only a field camera capable of night vision and no additional equipment. In particular, the hierarchical model integrates long-term dynamic patterns extracted from large-scale meteorological data with short-term dynamic patterns extracted from small-scale stream image data to jointly monitor water depth at a fine-level temporal resolution. In order to overcome the issue of limited availability of images, we introduce a transfer learning strategy and incorporate more accurate long-term patterns that enable the hierarchical model to perform competitively even with a small number of images. We evaluate our method on a real-world headwater stream monitoring dataset from the West Brook study area in western Massachusetts, United States. Our extensive experiments demonstrate that the hierarchical model outperforms several state-of-the-art methods for water depth monitoring, and that more accurate long-term patterns can better guide the monitoring of short-term patterns with excellent flexibility and less computational cost. The mean absolute error of our hierarchical model achieves a remarkable level of 4.9cm at the study site with 0.89m average water depths, and only 12.5cm at more drastically varied site with 3.95m average depths.

DenMerD: a feature enhanced approach to radar beam blockage correction with edge-cloud computing

AbstractIn the field of meteorology, the global radar network is indispensable for detecting weather phenomena and offering early warning services. Nevertheless, radar data frequently exhibit anomalies, including gaps and clutter, arising from atmospheric refraction, equipment malfunctions, and other factors, resulting in diminished data quality. Traditional radar blockage correction methods, such as employing approximate radial information interpolation and supplementing missing data, often fail to effectively exploit potential patterns in massive radar data, for the large volume of data precludes a thorough analysis and understanding of the inherent complex patterns and dependencies through simple interpolation or supplementation techniques. Fortunately, edge computing possesses certain data processing capabilities and cloud center boasts substantial computational power, which together can collaboratively offer timely computation and storage for the correction of radar beam blockage. To this end, an edge-cloud collaborative driven deep learning model named DenMerD is proposed in this paper, which includes dense connection module and merge distribution (MD) unit. Compared to existing models such as RC-FCN, DenseNet, and VGG, this model greatly improves key performance metrics, with 30.7$$\%$$ % improvement in Critical Success Index (CSI), 30.1$$\%$$ % improvement in Probability of Detection (POD), and 3.1$$\%$$ % improvement in False Alarm Rate (FAR). It also performs well in the Structure Similarity Index Measure (SSIM) metrics compared to its counterparts. These findings underscore the efficacy of the design in improving feature propagation and beam blockage accuracy, and also highlights the potential and value of mobile edge computing in processing large-scale meteorological data.

Climate change in the Catalan Pyrenees intersects with socioeconomic factors to shape crop diversity and management

Most studies on climate change’s impacts on agriculture focus on modeling techniques based on large-scale meteorological data, while few have investigated how farmer’s perception of climate change’s impacts can affect crop diversity and crop management practices, especially in industrialized contexts. To fill this gap, we conducted 24 semi-structured interviews in a study site located in the Catalan Pyrenees. Our results show for the first time in an industrialized context that farmers perceive multiple interrelated climate change impacts on local agroecosystems. For instance, snowfall and freeze events have decreased, which respondents associated with the increase of pests and diseases affecting both wild flora and cultivated plants. Similarly, changes in precipitation patterns lead to a perceived decrease in useful rain for agriculture. Farmers are also reporting changes in their management practices, such as increased irrigation or use of pesticides, which respond to these climatic factors but also to changes in the crops that are cultivated. Crop diversity is in decline in the area both at the species and landrace levels, especially in rainfed fields. This is mainly driven by socioeconomic factors such as agricultural abandonment or access to commercial seeds, although climate change factors such as increased pests or decreased rainfall can have an impact. Despite the crop diversity losses found, many landraces have been maintained, mainly due to their cultural value, and also new crop species have been introduced, which are now viable due to the increase in temperature. Although we focused on a specific case study, we found several trends that are also present in other contexts. Therefore, the results of this research are relevant at a global scale since they show that climate change is affecting mountain agroecosystems in industrialized contexts and may affect more drastically both agrobiodiversity and crop management practices in agroecosystems worldwide.

Improving accuracy of wind speed forecasting

Technique combines large-scale meteorological data with input from multiple modeling sites.

An interactive 4D spatio-temporal visualization system for hydrometeorological data in natural disasters

ABSTRACT Dynamic visualization of multidimensional hydrometeorological data is vital for decision-makers to catch situational awareness and command an emergency response in natural disasters. Nevertheless, few software tools can comprehensively visualize hydrometeorological data in different scales, dimensions, and time. In this paper, an interactive 4D spatio-temporal visualization system based on a virtual globe is proposed. Voxel-based data model and multi-level index are adopted to organize the field data in a unified data structure. Meanwhile, it is resampled in both spatial and temporal dimensions in memory to prepare smooth data stream for rendering. Ten field models, including large-scale volume rendering and adaptive streamline rendering, are accelerated and integrated to display the field data collaboratively. The profile analysis and eddy tracking improve user experience in interactively exploring specific scenes. The system is tested against both large-scale meteorological data in the atmosphere and small-scale hydrological data at the surface, using typhoon landfall and riverine flood, respectively. The results demonstrate the applicability and efficiency of the system to dynamically visualize hydrometeorological data.

An efficient dynamic volume rendering for large-scale meteorological data in a virtual globe

Texture slicing based volume rendering is widely used for data visualization, but dynamic high-fidelity rendering for large-scale meteorological data is still a challenge. The traditional octree-based multiresolution method increases the slicing and rendering batches, and it also has the visual defect of brick borders. This paper introduces an efficient dynamic volume rendering method for large-scale meteorological data in a virtual globe. First, the volumetric data is resampled and transformed adaptively according to the camera view. Second, the proxy geometry is calculated by slicing the whole spherical shaped volume along the ray from the camera to Earth center based on GPU. Finally, the rendering efficiency is optimized for interactive dynamic visualization. The proposed method is tested by visualizing Typhoon Rananim in an open source software, called World Wind. The experimental results demonstrate that the proposed method is more efficient and the visual effect is smoother than the octree-based multiresolution method.

Meteorological data analysis using MapReduce.

In the atmospheric science, the scale of meteorological data is massive and growing rapidly. K-means is a fast and available cluster algorithm which has been used in many fields. However, for the large-scale meteorological data, the traditional K-means algorithm is not capable enough to satisfy the actual application needs efficiently. This paper proposes an improved MK-means algorithm (MK-means) based on MapReduce according to characteristics of large meteorological datasets. The experimental results show that MK-means has more computing ability and scalability.

Skill in streamflow forecasts derived from large-scale estimates of soil moisture and snow

In the American west, operational forecasts for spring–summer streamflow rely heavily on snow-water storage. Simulations with a suite of land-modelling systems suggest that snow-water storage generally contributes most to forecasting skill, but that the contribution of soil moisture is often significant, too. Seasonal predictions of streamflow can benefit from knowledge of the amounts of snow and other water present in a basin when the forecast is issued1,2,3,4,5. In the American west, operational forecasts for spring–summer streamflow rely heavily on snow-water storage and are often issued at the time of maximum snow accumulation. However, forecasts issued earlier can also show skill, particularly if proxy information for soil moisture, such as antecedent rainfall, is also used as a predictor1,4. Studies using multiple regression approaches and/or model-produced streamflows6,7,8,9 indeed suggest that information on soil moisture—a relatively underappreciated predictor—can improve streamflow predictions. Here, we quantify the relative contributions of early-season snow and soil moisture information to the skill of streamflow forecasts more directly and comprehensively: in a suite of land-modelling systems, we use the snow and soil moisture information both together and separately to derive seasonal forecasts. Our skill analysis reveals that early-season snow-water storage generally contributes most to skill, but the contribution of early-season soil moisture is often significant. In addition, we conclude that present-generation macroscale land-surface models forced with large-scale meteorological data can produce estimates of water storage in soils and as snow that are useful for basin-scale prediction.

Construction of consistent ice core accumulation time series from large-scale meteorological data: development and description of a regression model for one North Greenland ice core

CR Climate Research Contact the journal Facebook Twitter RSS Mailing List Subscribe to our mailing list via Mailchimp HomeLatest VolumeAbout the JournalEditorsSpecials CR 20:141-151 (2002) - doi:10.3354/cr020141 Construction of consistent ice core accumulation time series from large-scale meteorological data: development and description of a regression model for one North Greenland ice core Traute Crüger1,*, Hans von Storch2 1Institute of Meteorology, University of Hamburg, Bundesstrasse 55, 20146 Hamburg, Germany 2GKSS Research Centre, Max-Planck-Strasse, 21502 Geesthacht, Germany *E-mail: crueger@dkrz.de ABSTRACT: An empirical downscaling relationship is constructed which consistently estimates annual ice accumulation from large-scale meteorological data. The method by which the relationship was developed and the application of the method to one ice core are described. The statistical technique is based on a stepwise multiple linear regression using yearly accumulation as the predictand and the principle components (PCs) of seasonally averaged large-scale atmospheric fields as predictors. In order to separate the accumulation variations related to circulation and thermodynamics, the first step involves the stream-function field as the predictor. In the second step, temperature PCs are related to the residuals between real ice accumulation and accumulation described by stream-function PCs. One model is fitted for a North Greenland ice core. The atmospheric data are monthly NCEP Reanalysis data from 1948 to 1992. A statistical relationship is found which reproduces about 71.5% of local ice accumulation variability. The relationship involves 3 physically plausible stream-function patterns, representing a seasonal mean (May to August) over Greenland, the North Atlantic and North Europe, which describes 64% of variance. In the second step, an additional contribution of the temperature field to the explained variance of 7.5% is achieved. The temperature PCs represent the annual mean 700 hPa pattern covering the area from east Canada to east Greenland. KEY WORDS: Ice core accumulation · Downscaling · NCEP-Reanalysis data Full text in pdf format PreviousNextExport citation RSS - Facebook - Tweet - linkedIn Cited by Published in CR Vol. 20, No. 2. Online publication date: February 21, 2002 Print ISSN: 0936-577X; Online ISSN: 1616-1572 Copyright © 2002 Inter-Research.

An Improved Statistical-Dynamical Downscaling Scheme and its Application to the Alpine Precipitation Climatology

An improved statistical-dynamical downscaling method for the regionalization of large-scale climate analyses or simulations is introduced. The method is based on the disaggregation of a multi-year time-series of large-scale meteorological data into multi-day episodes of quasi-stationary circulation. The episodes are subsequently grouped into a defined number of classes. A regional model is used to simulate the evolution of weather during the most typical episode of each class. These simulations consider the effects of the regional topography. Finally, the regional model results are statistically weighted with the climatological frequencies of the respective circulation classes in order to provide regional climate patterns.

Statistical downscaling of monthly mean air temperature to the beginning of flowering of Galanthus nivalis L. in Northern Germany

We have examined the relationship between phenological data and concurrent large-scale meterological data. As phenological data we have chosen the beginning of the flowering of Galanthus nivalis L. (flowering date) in Northern Germany, and as large-scale meteorological data we use monthly mean near-surface air temperatures for January, February and March. By means of canonical correlation analysis (CCA), a strong linear correlation between both sets of variables is identified. Twenty years of observed data are used to build the statistical model. To validate the derived relationship, the flowering date is downscaled from air temperature observations of an independent period. The statistical model is found to reproduce the observed flowering dates well, both in terms of variability as well as amplitude. Air temperature data from a general circulation model of climate change are used to estimate the flowering date in the case of increasing atmospheric carbon dioxide concentration. We found that at a time of doubled CO2 concentration (expected by about 2035) G. nivalis L. in Northern Germany will flower ∼2 weeks and at the time of tripled CO2 concentration (expected by about 2085) ∼4 weeks earlier than presently.