In this post, we are going to see how to plot two different Sentinel-5P products on the same geographic aerea.

The products we are going to plot are the cloud product from the VIIRS instrument on Suomi-NPP, regridded on the TROPOMI grid, and the CH4 product from TROPOMI.

The CH4 measurements must be acquired on land and cloud free pixels only, and this is something we can visually check by superimposing the two plots.

To format our plot, we are going to make use of Python, xarray, cartopy, and matplotlib.

On February 19, 2018, Indonesia’s Mount Sinabug, a stratovolcano on the island of Sumatra, erupted violently, spewing ash at least 5 to 7 kilometers into the air over Indonesia.

The erupting lava dome obliterated a chunk of the peak as it erupted. Plumes of hot gas and ash rode down the volcano’s summit and spread out in a 5-kilometer diameter, coating surrounding villages in ash.

In this post, we are going to see how this event has been recorded by the TROPOMI instrument on board of ESA‘s Sentinel-5P earth observation satellite, using Python, xarray, and cartopy

I recently started to look into the R language, as part of my growing interest in statistical learning and data analysis. As a first impression, I felt a bit overwhelmed by the number of differences with Python, the vast amount of oddly named libraries, and the many different ways to address a given problem, as opposed to Python, where there is usually a general consensus about a pythonic way to do things.

Anyway, one of the most effective ways to learn a new programming language is to try using it to solve familiar problems. And that is precisely what I have been doing in the past few days, looking into a fast and efficient way to plot some fresh Sentinel-5P NO₂ data over a world map.

In this post, we are going to read the contents of several NetCDF4 products and turn them into a single data frame, where each row will contain an observation and the related geodetic coordinates. We are then going to slice the global data frame over a region of interest, and plot the correspondent data over a map with ggplot2. I am sure there are more clever and elegant ways to approach this task and I will make sure to document them in future posts.

The Sentinel-5P data I work on lately is typically organised on an irregular two-dimensional grid whose dimensions are scanline (along track dimension) and ground pixel (across track dimension). Latitude and longitude information for each ground pixel are stored in auxiliary coordinate variables.

I have recently stumbled upon the problem of locating the closest ground pixel to a reference point, identified by its geodetic coordinates.

In this post, we are going to define an algorithm to solve this problem by using coordinate transformations, $k$-dimensional trees, and xarray pointwise indexing.

As part of the commissioning phase of the TROPOMI instrument on board of the ESA’s Sentinel-5P spacecraft, a dedicated radiometric measurement was put in place to estimate radiometric errors in Level 1B products based on S5P early in-flight calibration measurements.

Simulations of atmospheric transmission spectra showed that the atmosphere is opaque at tangent altitudes below 15-10km, near the O₂ A Band at 761 nm.
Measurements in this range, taken in a so-called solar occultation geometry, could potentially be used to detect residual out-of-band stray light effects.

In this post, we are going to show how to use Matplotlib to display an animated plot of the solar irradiance spectra detected during this dedicated solar calibration measurement, as a function of time – which roughly translates in decreasing limb tangent heights — and observe the strong O₂ A Band absorption features occurring around 761 nm.

In the early phase of the Sentinel-5P commissioning, a number of pitch rotation manoeuvres were put in place, in order to perform specific in-flight calibration measurements.

In this post we are going to try to outline a generalised way in Python to identify the pitch rotations occurring during the set of planned spacecraft manoeuvres. This will allow us to derive manoeuvres execution times and their duration.

The process under analysis can be considered a special case of a much broader statistical and numerical processing class of problems, called step detection, that is the process of finding abrupt changes (steps, jumps, shifts) in the mean level of a time series or signal.