About one German paper on AirHES
A long time ago, in December 2016, someone sent me a paper by a German student (Kilian Heilgemair) called "Theoretische Ertragsberechnung von fliegenden Nebelfängern anhand von Satellitendaten des MERRA-Projekts" ("Theoretical calculation of yields of flying fog catchers from satellite data of the MERRA project").
It literally quoted pages from my site (e.g., page 5), but there was no mention of me, the site, or AirHES itself. However, I didn't care about it, although it was strange.
Nevertheless, the paper was written quite interestingly. The method of theoretical calculation of the amount of moisture in the air (absolute humidity) based on the Magnus formula and relative humidity data was used, as well as calculation of altitude based on the barametric formula and pressure data at a given level. However, in the formulas on page 8 the dimensions were confused: first absolute humidity was expressed in g/m3 instead of kg/m3, and then water yield from the mesh was expressed in m3/s instead of kg/s.
In fact, these calculations were unnecessary because the MERRA project and MERRA-2 already contain absolute humidity and altitude (QV, H) data that can be obtained directly and recalculated based on this methodology.
MERRA project specification for the AI3CPASM dataset
The calculations of mesh efficiency practically repeated the well-known work of scientists from MIT, the comparative analysis of which I also used in my paper on mesh optimization. In addition, data from NASA Aqua and Terra satellites were used there to obtain the average size of water droplets, although this parameter practically has little influence on the efficiency calculation, but, to my surprise, data on cloud water content were not used.
So, I decided to cross-check these calculations in two directions at once: with observed cloud (liquid) water content (LWC) data from NASA CloudSat and with the calculated data from the MERRA-2 project. For comparison, I took one of the examples in this paper, namely the calculations for Lisbon in 2014.
The CloudSat satellite passed over Lisbon (within 1 degree latitude and longitude) 26 times during 2014 and recorded LWC data, which are presented in the following table.
CloudSat (Lisbon, 2014) data on cloud (liquid) water content, droplet flux and specific hydropower by altitude
Water content data was obtained directly from the satellite, but droplet flux and specific hydropower was calculated with wind speeds borrowed from the MERRA-2 project, whose data are presented in the table below.
MERRA-2 (Lisbon, 2014) data on cloud (liquid) water content, droplet flux and specific hydropower by altitude
Further on, the issues begin, which I have already tried to point out in the last publication on the precipitation paradox. The author of the paper, describing on p.25 the program and calculation methodology, states that LWC calculation is based on relative humidity (RH) and temperature (T) values, i.e. on the Magnus formula (pp.7-8). Following him, I recalculate the absolute humidity (SH) both directly from the QV data and by his methodology (via RH and T) and get almost a perfect match (the error does not exceed 1%).
Comparison of absolute humidity (SH) calculations by QV and RH,T
The values of absolute humidity SH (whichever way they are calculated) are roughly consistent with the amount of precipitation. For example, taking the average lifetime of moisture in the atmosphere as 10 days, we obtain an annual precipitation of 807.85 liters per m2 (which is equal to the height in mm, since each liter is equal to 1 mm * m2), whereas the statistics for Lisbon give about [587] mm.
However, this value of "avg LWC SH" is (alas!) 2-3 orders of magnitude larger than the real "avg LWC", which can be obtained either from CloudSat satellite data or from MERRA-2 computational data using the QL parameter, which the author does not use anywhere. Nevertheless, when describing the results of the calculations (p.28 and beyond), he unexpectedly gives values very close to the values calculated for the real LWC. The question remains, where did he get them from?
Let us compare, for example, his monthly rainfall data (p.31) for Lisbon in 2014 with my normalized calculations (green circles). For comparison, I picked a mesh efficiency value of 14.5% to bring the average to about the same value of ~300 l/m2 per month.
Comparison of monthly data for Lisbon in 2014 in the German paper and in my calculations (green circles)
You can see that the data of my calculations are very similar to the data of the German paper, although the absurdity starts with the fact that the German author compares these his calculations with precipitation without noticing that they are completely different m2 - precipitation = horizontal m2, and on a mesh = vertical m2, i.e. on one horizontal m2 you can place upwards of thousands of vertical m2 and get any amount of water.
A more or less correct comparison can be obtained only through economic criteria. For example, let's calculate the TE parameters for AirHES under full optimization using my complex model for Lisbon in 2014. If we use CloudSat data to calculate the maximum possible cloud water content vertically (i.e., constantly adjusting the elevation of the AirHES mesh), we obtain a value of 98.52662 mg/m3 (global max), which, when optimized in the model, gives, for example, the following figures:
I.e. payback period ~ 1.2 years, EE price ~ 9.9 cents/kWh, water price ~ 13.7 cents/m3, productivity ~ 4.86 liters per day per 1 physical m2 of single-layer mesh (Chilean type). The calculation was made for a 100 by 100 m mesh with a lifetime of up to 10 years, i.e. a total water flow of about 50 tons per day.