When electron density measurements disagree

Figure: The two panels show two independent electron density measurements made during the two flights of the MAXIDUSTY campaign in the summer of 2016. 'Far' denotes Faraday rotation while 'mNLP' denotes multi Needle Langmuir Probe. The disagreement is clear.

The electron density is arguably one of the most important parameters readily derived from any dataset obtained during campaigns studying different layers of the ionosphere. Radars are cool (citation needed) in that they can very often provide good measurements of the electron density, especially in the E- and F-region of the ionosphere. The height resolution is limited, but radars can stay operative for long periods of time. In the D-region, however, things get freaky (citation needed). Most radars cannot provide good electron density measurements here, and the best option for absolute measurements is in-situ probing.

The figure on top of the article shows two observations of the electron density, done with two different means of measurement, from the summer of 2016 during the MAXIDUSTY rocket campaign. The two methods are Faraday rotation and Langmuir probes in the form of needle probes (mNLP). They clearly disagree; above ~95 km the difference is not worse than a short discussion about rocket payload sheath/aerodynamical effects, however, around 80 km the difference is around an order of magnitude.

The region between 80 and 90 km (around the mesopause) is particularly interesting since it is the coldest environment in the Earth's atmosphere during the summer months. The cold temperatures, reaching as low as 100 K (!), makes it possible for a range of ice and dust clouds to exist. The ice and dust in this height region constitutes, together with electrons and ions, a plasma. This special kind of plasma with free charged nanoscale particles is often called a dusty or complex plasma.

How is the digression above important? All plasmas are more or less quasi-neutral (read 1=1), which means that to really know something about the ice and dust particles you need to know about the electrons. Therefore, for mesospheric rocket experiments there is always an onboard instrument to derive the absolute electron density. For some decades now, the main method, and often the only onboard method to measure the electrons, has been Faraday rotation. This was proposed in the late 60's by Martin Friedrich et al. (now professor at TU Graz). The measurements from Faraday rotation more or less agree with radar measurements, but has a limitation in its height resolution. In addition to this method, there has been an ongoing effort to to make rocket-borne Langmuir probes a preferred option. The advantage with such probes is that they do not require on ground instrumentation (as Faraday rotation requires) and have superior height resolution, but they are on the other hand more difficult to calibrate. 

During the MAXIDUSTY campaign, lead by PI Ove Havnes at UiT, we did a number of measurements which would greatly benefit from accurate electron density measurements. Just to be sure, we even included two independent means of measurement; the two mentioned instruments. To our great surprise, the two instruments disagreed strongly in our main region of interest, the mesopause. As a researcher, how would you continue? Since the instruments were not your own, some people would maybe be tempted to leave the discussion to the owners of the probes. Most people would however be intrigued and baffled by the result, and try to figure out what is happening.

So what do you think about the discrepancy? As a hint, I could say that the SAURA MF radar agrees more with the Faraday rotation measurements. Langmuir probes are basically just biased pieces of metal, and therefore a number of issues can arise in their calibration. The mNLP-probes are very thin probes (smaller than 1 mm in thickness), designed to be smaller than all other governing length scales (debye length, mean free path) so that orbital motion theory can be applied. The mean free path is still exactly that: mean. My guess is therefore that the probe sheath is not truly collisionless, which may explain a part of the difference.