Sea surface temperature (SST) plays an important role in a number of ecological processes and can vary over a wide range of time scales, from daily to decadal changes. SST influences primary production, species migration patterns, and coral health. If temperatures are anomalously warm for extended periods of time, drastic changes in the surrounding ecosystem can result, including harmful effects such as coral bleaching. This layer represents the standard deviation of SST (degrees Celsius) of the weekly time series from 2000-2013. Three SST datasets were combined to provide continuous coverage from 1985-2013. The concatenation applies bias adjustment derived from linear regression to the overlap periods of datasets, with the final representation matching the 0.05-degree (~5-km) near real-time SST product. First, a weekly composite, gap-filled SST dataset from the NOAA Pathfinder v5.2 SST 1/24-degree (~4-km), daily dataset (a NOAA Climate Data Record) for each location was produced following Heron et al. (2010) for January 1985 to December 2012. Next, weekly composite SST data from the NOAA/NESDIS/STAR Blended SST 0.1-degree (~11-km), daily dataset was produced for February 2009 to October 2013. Finally, a weekly composite SST dataset from the NOAA/NESDIS/STAR Blended SST 0.05-degree (~5-km), daily dataset was produced for March 2012 to December 2013. The standard deviation of the long-term mean SST was calculated by taking the standard deviation over all weekly data from 2000-2013 for each pixel.

📰 Related Paper

Sichkar V. N. Effect of various dimension convolutional layer filters on traffic sign classification accuracy. Scientific and Technical Journal of Information Technologies, Mechanics and Optics, 2019, vol. 19, no. 3, pp. DOI: 10.17586/2226-1494-2019-19-3-546-552 (Full-text available here ResearchGate.net/profile/Valentyn_Sichkar)

Test online with custom Traffic Sign here: https://valentynsichkar.name/mnist.html

:mortar_board: Related course for classification tasks

Design, Train & Test deep CNN for Image Classification. Join the course & enjoy new opportunities to get deep learning skills: https://www.udemy.com/course/convolutional-neural-networks-for-image-classification/

https://github.com/sichkar-valentyn/1-million-images-for-Traffic-Signs-Classification-tasks/blob/main/images/slideshow_classification.gif?raw=true%20=470x516" alt="CNN Course" title="CNN Course">

🗺️ Concept Map of the Course

https://github.com/sichkar-valentyn/1-million-images-for-Traffic-Signs-Classification-tasks/blob/main/images/concept_map.png?raw=true%20=570x410" alt="Concept map" title="Concept map">

👉 Join the Course

https://www.udemy.com/course/convolutional-neural-networks-for-image-classification/

Content

This is ready to use preprocessed data saved into pickle file.
Preprocessing stages are as follows:
- Normalizing whole data by dividing / 255.0.
- Dividing whole data into three datasets: train, validation and test.
- Normalizing whole data by subtracting mean image and dividing by standard deviation.
- Transposing every dataset to make channels come first.

mean image and standard deviation were calculated from train dataset and applied to all datasets.
When using user's image for classification, it has to be preprocessed firstly in the same way: normalized, subtracted with mean image and divided by standard deviation.

Data written as dictionary with following keys:
x_train: (59000, 1, 28, 28)
y_train: (59000,)
x_validation: (1000, 1, 28, 28)
y_validation: (1000,)
x_test: (1000, 1, 28, 28)
y_test: (1000,)

Contains pretrained weights model_params_ConvNet1.pickle for the model with following architecture:
Input --> Conv --> ReLU --> Pool --> Affine --> ReLU --> Affine --> Softmax

Parameters:

• Input is 1-channeled GrayScale image.
• 32 filters of Convolutional Layer.
• Stride for Pool is 2 and height = width = 2.
• Number of hidden neurons is 500.
• Number of output neurons is 10.

Architecture also can be understood as follows:
https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F3400968%2Fc23041248e82134b7d43ed94307b720e%2FModel_1_Architecture_MNIST.png?generation=1563654250901965&alt=media" alt="">

Acknowledgements

Initial data is MNIST that was collected by Yann LeCun, Corinna Cortes, Christopher J.C. Burges.

An Open Context "predicates" dataset item. Open Context publishes structured data as granular, URL identified Web resources. This "Variables" record is part of the "Avkat Archaeological Project" data publication.

standard deviation of 12D measured via Incubation in mg C/m^3. Part of dataset Gradients 1-KOK1606 - Net Primary Productivity (via 14C method)

Version 1 is the current version of the dataset.This collection MODFDS_SDV_GLB_L3 provides level 3 standard deviation of climatological monthly frequency of dust storms (FDS) over land from 175°W to 175°E and 80°S to 80°N at a spatial resolution of 0.1˚ x 0.1˚. It is derived from Level 2, the Moderate Resolution Imaging Spectroradiometer (MODIS) Deep Blue aerosol products Collection 6.1 from Terra (MOD04_L2). The dataset is the standard deviation of climatological monthly mean for each month over 2000 to 2022.The FDS is calculated as the number of days per month when the daily dust optical depth is greater than a threshold optical depth (e.g., 0.025) with two quality flags: the lowest (1) and highest (3). It is advised to use flag 1, which is of lower quality, over dust source regions, and flag 3 over remote areas or polluted regions. Eight thresholds (0.025, 0.05, 0.1, 0.25, 0.5, 0.75, 1, 2) are saved separately in eight files.If you have any questions, please read the README document first and post your question to the NASA Earthdata Forum (forum.earthdata.nasa.gov) or email the GES DISC Help Desk (gsfc-dl-help-disc@mail.nasa.gov).

Prior research has shown that sediment budgets, and therefore stability, of microtidal marsh complexes scale with areal unvegetated to vegetated marsh ratios (UVVR) suggesting these metrics are broadly applicable indicators of microtidal marsh vulnerability. This effort has developed the UVVR metric using Landsat 8 satellite imagery for the coastal areas of the contiguous United States (CONUS). These datasets provide annual averages of 1) developed, 2) vegetated, 3) unvegetated fractional covers and 4) an unvegetated to vegetated ratio (UVVR) at 30-meter resolution over the coastal areas of the contiguous United States for the years 2014-2018. Additionally, multi-year average values of vegetated fractional cover and its standard deviation are provided for the coastal wetlands of the contiguous United States based on the National Wetland Inventory delineation. Finally, a UVVR based on the annually-averaged vegetated fractional cover is also provided for the same extent.

AD8232

Means, standard deviations, and sample sizes for Letter-Word Identification standard scores for the target sample.

propagated standard deviation of 24SL-D measured via Incubation in mg C/m^3. Part of dataset Gradients 1-KOK1606 - Net Primary Productivity (via 14C method)

Brazil Market Expectation: Regulated Prices: Next Calendar Year: Standard Deviation data was reported at 0.430 % in Jun 2019. This records a decrease from the previous number of 0.510 % for May 2019. Brazil Market Expectation: Regulated Prices: Next Calendar Year: Standard Deviation data is updated monthly, averaging 0.615 % from May 2003 (Median) to Jun 2019, with 194 observations. The data reached an all-time high of 1.430 % in Nov 2003 and a record low of 0.250 % in Jan 2018. Brazil Market Expectation: Regulated Prices: Next Calendar Year: Standard Deviation data remains active status in CEIC and is reported by Central Bank of Brazil. The data is categorized under Brazil Premium Database’s Business and Economic Survey – Table BR.SB037: Market Expectation: Regulated Prices. Market Expectations System was implemented in November 2001, previous projections were collected from incipient through telephone contacts, transcribed into spreadsheets and consolidated manually. Some empty time points occurred because the Market didn´t have the expectation for those days. Prices administered by contract and monitored Prices administered by contract and monitored are those whose sensitivity to factors of supply and demand is lower, which does not necessarily imply that they are directly regulated by the government.

Chlorophyll-a is a widely used proxy for phytoplankton biomass and an indicator for changes in phytoplankton production. As an essential source of energy in the marine environment, the extent and availability of phytoplankton biomass can be highly influential for fisheries production and dictate trophic structure in marine ecosystems. Changes in phytoplankton biomass are predominantly effected by changes in nutrient availability, through either natural (e.g., turbulent ocean mixing) or anthropogenic (e.g., agricultural runoff) processes. This layer represents the standard deviation of the 8-day time series of chlorophyll-a (mg/m3) from 2002-2013. Monthly and 8-day 4-km (0.0417-degree) spatial resolution data were obtained from the MODIS (Moderate-resolution Imaging Spectroradiometer) Aqua satellite instrument from the NASA OceanColor website (http://oceancolor.gsfc.nasa.gov). The standard deviation was calculated over all 8-day chlorophyll-a data from 2002-2013 for each pixel. A quality control mask was applied to remove spurious data associated with shallow water, following Gove et al., 2013. Nearshore map pixels with no data were filled with values from the nearest neighboring valid offshore pixel by using a grid of points and the Near Analysis tool in ArcGIS then converting points to raster.

Chlorophyll-a, is a widely used proxy for phytoplankton biomass and an indicator for changes in phytoplankton production. As an essential source of energy in the marine environment, the extent and availability of phytoplankton biomass can be highly influential for fisheries production and dictate trophic structure in marine ecosystems. Changes in phytoplankton biomass are predominantly effected by changes in nutrient availability, through either natural (e.g., turbulent ocean mixing) or anthropogenic (e.g., agricultural runoff) processes. This layer represents the standard deviation of the 8-day time series of chlorophyll-a (mg/m3) from 1998-2018. Data products generated by the Ocean Colour component of the European Space Agency (ESA) Climate Change Initiative (CCI) project. These files are 8-day 4-km composites of merged sensor products: Global Area Coverage (GAC), Local Area Coverage (LAC), MEdium Resolution Imaging Spectrometer (MERIS), Moderate Resolution Imaging Spectroradiometer (MODIS) Aqua, Ocean and Land Colour Instrument (OLCI), Sea-viewing Wide Field-of-view Sensor (SeaWiFS), and Visible Infrared Imaging Radiometer Suite (VIIRS). The standard deviation was calculated over all 8-day chlorophyll-a data from 1998-2018 for each pixel. A quality control mask was applied to remove spurious data associated with shallow water, following Gove et al., 2013. Nearshore map pixels with no data were filled with values from the nearest neighboring valid offshore pixel by using a grid of points and the Near Analysis tool in ArcGIS then converting points to raster. Data source: https://oceanwatch.pifsc.noaa.gov/erddap/griddap/esa-cci-chla-8d-v5-0.graph

Standard deviation of responses for 'Life Satisfaction' in the First ONS Annual Experimental Subjective Wellbeing survey.

Means and standard deviations of intercepts and slopes of Letter-Word Identification W and standard scores for the target sample.

This part of the data release contains a grid of standard deviations of bathymetric soundings within each 0.5 m x 0.5 m grid cell. The bathymetry was collected on February 1, 2011, in the Sacramento River from the confluence of the Feather River to Knights Landing. The standard deviations represent one component of bathymetric uncertainty in the final digital elevation model (DEM), which is also available in this data release. The bathymetry data were collected by the USGS Pacific Coastal and Marine Science Center (PCMSC) team with collaboration and funding from the U.S. Army Corps of Engineers. This project used interferometric sidescan sonar to characterize the riverbed and channel banks along a 12 mile reach of the Sacramento River near the town of Knights Landing, California (River Mile 79 through River Mile 91) to aid in the understanding of fish response to the creation of safe habitat associated with levee restoration efforts in two 1.5 mile reaches of the Sacramento River between River Mile 80 and 86.

The differential solubility of ferromanganese oxides can lead to stratigraphic separation of iron and manganese. Results of chemical analysis of a sequence of ferromanganese nodules overlying iron-rich crusts in northern Green Bay show that selec¬tive ion transport is important in concentrating manganese and associated trace elements near the oxygenated water-sediment interface. Manganese carbonate, which cements ferromanganese nodules, occurs in dark-gray silty sands that are located adjacent to the organic-rich muds of southern Green Bay. These muds contain an average of approximately 3.5 ppm (6x10-5M) interstitial Mn with 2.8 meq/l carbonate alkalinity. Thermodynamic calculation shows that interstitial water approaches equilibrium with MnCO3 in the upper 10 cm of sediment. This carbonate has a composition (Mn73Ca22Fe5)CO3 and has been identified as rhodochrosite.

Techical Information: Transitions interpolated within coring or sampling gaps can be identified from the associated large range of uncertainty. Boundaries of thin or otherwise poorly defined polarity intervals are indicated by a query.

Contents of rare earth elements (REE) in standard samples of Fe-Mn nodules (SDO-5, 6), Fe-Mn crust (SDO-7), and red clay (SDO-9) have been determined by ICP-MS and instrumental neutron activation analysis. Reproducibility of ICP-MS was 5-6%. These results are discussed and compared with other data. It has been found that distribution of REE in the standard samples of ocean Fe-Mn ores and red clay is highly homogenous.

Mean reaction time (RTs; in ms) followed by the Standard Error of the Mean (SEM) and mean proportion of errors (ERR; in %) followed by standard deviation (SD), depicted separately for same and switched additions (+) and subtractions (−), for the original symbol-switching task (from Experiment 1), the symbol-switching task with letters (from Experiment 2), and the stimulus offset condition from Experiment 2, where the letters offset upon response (Fast Offset).

The mean (), standard deviation (), and the ratio () of the recognition rates (%) for MR_2DLDA when different face images are used as the training samples.