Google Satellite Maps Downloader 6.55 Serial Number
Pamula Harrison
In this example some of the basic functionalities of the ESA Atmospheric Toolbox to handle TROPOMI data are demonstrated. This case focuses on the use of toolbox's HARP component in Python, implemented as a Jupyter notebook.
The ESA Copernicus TROPOMI instrument onboard Sentinel 5p satellite observes atmospheric constituents at very high spatial resolution. In this tutorial we will demonstrate basic data reading and plotting procedures using TROPOMI SO2 observations. We use observations that were obtained during the explosive eruption of La Soufriere volcano in the Caribbean in April 2021. The eruption released large amounts of SO2 into the atmosphere, resulting extensive volcanic SO2 plumes that were transported long distances. This notebook will demonstrate how this event can be visualized using TROPOMI SO2 observations and HARP.
The instructions on how to get started with the Atmospheric toolbox using Python and install HARP can be found here (add link to getting started jupyter notebook). Please note that if you have installed HARP in some specific python environment, check that you have activated the environment before running the script.
The second step is to import the TROPOMI Level 2 SO2 file using harp.import_product(). If the file does not yet exist on your local machine, we use the avl library to automatically download the file from the Copernicus Dataspace Ecosystem (CDSE). (Because the original netcdf file is large, both downloading and importing the file might take a while.)
We use eofetch to download the S5P product. To be able to perform the download yourself you will need to retrieve and configure credentials as described in the eofetch README.Alternatively, you can download the file manually and put it in the same directory as this notebook.
After a successful import, you have created a python variable called product. The variable product contains a record of the SO2 product variables, the data is imported as numpy arrays. You can view the contents of product using the Python print() function:
From the listing above you see e.g. that the unit of the SO2_column_number_density variable is mol/m^2. Type of the product and the shape (size) of the SO2_column_number_density data array can be checked with the following commands:
Here it is important to notice that harp.import_product command imports and converts the TROPOMI Level 2 data to a structure that is compatible with the HARP conventions. This HARP compatible structure is different from the netcdf file structure. This HARP conversion includes e.g. restructuring data dimensions or renaming variables. For example, from the print commands above it is shown that after HARP import the dimension of the SO2_column_number_density data is time (=1877400), whereas working with netcdf-files directly using e.g. a library such as netCDF4, the dimensions of the same data field would be a 2D array, having Lat x Lon dimension.
HARP has builtin converters for a lot of atmospheric data products. For each conversion the HARP documentation contains a description of the variables it will return and how it mapped them from the original product format. The description for the TROPOMI SO2 product can be found here.
HARP does this conversion such that data from other satellite data products, such as OMI, or GOME-2, will end up having the same structure and naming conventions. This makes it a lot easier for users to deal with data coming from different satellite instruments.
Now that the TROPOMI SO2 data product is imported, the data will be visualized on a map. The parameter we want to plot is the "SO2_column_number_density", which gives the total atmospheric SO2 column. For this we will be using cartopy and the scatter function. This plotting function is based on using only the pixel center coordinates of the satellite data, and not the actual latitude and longitude bounds. The scatter function will plot each satellite pixel as coloured single dot on a map based on their lat and lon coordinates. Cartopy also provides other plotting options, such as pcolormesh. However, in pcolormesh the input data needs to be a 2D array. This type of plotting will be demonstrated in the another use cases.
First, the SO2, latitude and longitude center data are defined. In addition, units and description of the SO2 data are read that are needed for the colorbar label. For plotting a colormap named "batlow" is chosen from the cmcrameri library. The cmcrameri provides scientific colormaps where the colour combinations are readable both by colour-vision deficient and colour-blind people. The Crameri colormap options can be viewed here. In the script the colormaps are called e.g. as cm.batlow. If you wish to use reversed colormap, append _r to the colormaps name. With vmin and vmax the scaling of the colormap values are defined.
Next the figure properties will be defined. By using matplotlib figsize argument the figure size can be defined, plt.axes(projection=ccrs.PlateCarree()) sets a up GeoAxes instance, and ax.coastlines() adds the coastlines to the map. The actual data is plotted with plt.scatter command, where lat and lon coordinates are given as input, and the dots are coloured according to the pixel SO2 value (SO2val).
Globally, the need for groundwater (GW) increases as population in a region increases. Recently, excessive quantity or volume of groundwater has been dug to meet water demand of this increasing population (Mahamat et al. 2020). It is considered as a valuable natural resource for agriculture in different communities, and it is relatively healthy for human consumption compared to other hydrological structures (especially surface water) since its exposure to pollution is less (Oke and Fourier 2017; Naghibi et al. 2016). Unfortunately, over $250 billion is spent globally on an annual basis due to short supply of sanitation and healthy drinking water, hence, the need to explore the potential availability of groundwater in a cost-effective manner.
Conventionally, different disciplines have attempted the delineation of GW potential zones. In geophysics, diverse surface geophysical techniques have been used in exploring groundwater including gravity, seismic refractive, radioactivity, magnetic, electromagnetic and electrical resistivity methods (Theophilus et al. 2018). These methods often require very deep penetration of geological materials for the purpose of examining these materials which are indicators of potential groundwater. Among these techniques, the electromagnetic and electrical resistivity methods have been reported to be the most utilized globally because they are capable of burrowing deeper into harsh rock terrains and have proven to be very effective when compared to other geophysical techniques (Oladejo et al. 2015; Mohameden and Ehab 2017). However, these techniques are still limited as they require rigorous field works, sensitive and expensive equipment, cumbersome analysis of geology and tectonic situations on a large scale, a detailed understanding of aquifer types, and in some cases, they require wide and deep puncturing of ground surfaces, leaving the area of interest vulnerable to environmental hazards and a deteriorated public health (Shadrach and Osazee 2020). Added to the weaknesses of conventional approach of GW exploration is the necessity of integrating two or more of these techniques before optimal results can be achieved (Okpoli and Ozomoge 2020). Remarkably, in developing countries like most African Countries, mapping of GW takes a limitless amount of time, and additional efforts traceable to lack of required human and financial resources. Also, the success rate of conventional approaches is a pointer to its inefficiency. Diaz-Alcaide et al. (2017) identified some regions in Mali where GW exploration (through borehole drilling) success rate falls below 40%. Muchingami et al. (2019) likewise reported a 25% success rate in crystalline areas of Zimbabwe. Therefore, there is need to explore a more sophisticated, timely and efficient method of delineating GW potential zones so as to meet the rising portable water need of a rising global population.
Conversely, remote sensing (RS) has become a useful technique of providing overviews of water cycle components on a large scale. Current developments of RS technologies (its sensor and processing techniques) have made significant advances in spatial and temporal resolution. Observations and models from RS techniques are useful and efficient resources for monitoring and managing GW resources. Thematic maps of factors influencing the retention and movement of GW (such as slope, rainfall, land use/land cover, drainage density, and lineament) are derivable from remotely sensed data and are best analysed using geographical information system (GIS).
For over two decades, numerous scientific studies have combined or integrated RS and GIS technologies, exploiting these factors controlling GW existence, for understanding and defining GW potential zones in a region (Chowdhury et al. 2009; Zoheir and Emam 2014; Al-Djazouli et al. 2019; Jadhav et al. 2022). Findings from these scientific studies show that the results obtained were satisfactory. Gyeltshen et al. (2020) investigated the possibility of improving the output of geospatial technology by integrating the techniques with geophysical approach. While the geophysical survey conducted only served the purpose of categorizing aquifers in the study area, the emphasis was mainly on the geospatial techniques.
Also, several researchers have successfully embraced geospatial technologies over the last two decades for the delineation of potential GW regions (Solomon and Quiel 2009, Chowdhury et al. 2009; Elbeih 2015; Agarwal and Garg 2016). However, a proper ranking and hierarchical combination of GW indicators will enhance and increase the reliability of geospatial approach of mapping and zoning GW potential areas. According to Machiwal et al. (2011), multi-criteria decision-making will furnish an operative system for managing water by adding rigour, transparency and framework decisions. Therefore, to optimize the result-oriented and cohesive system that RS and GIS provide for managing and combining different variables that influence GW availability and movement, there is need for an integration of these tools using multi-criteria decision analysis (MCDA) (Abrams et al. 2018; Altafi et al. 2020). Remarkably, previous studies have combined two or all of RS, GIS and MCDA for mapping and delineating GW potential zones in different regions, and outputs from these researches reveal that they remain as robust as possible, with much more benefits (including simplicity, transparency, and reliability) than the automatic (machine) learning approaches (Abrams et al. 2018; Ejepu et al. 2022).
7fc3f7cf58