15 Feb

HAMP sneaks in the queue.

First, I would like to reply to Bernhard Mayer that not all about HAMP is a mysterious zig-zag sawtooth ;).  Actually, the hardware issues, we were facing two and a half years ago, were mostly about one of three modules.  The remaining two modules, however, provide enough data to have an interesting time.

Before reminding you how clouds looked like during NARVAL2, I would like to give a short reminder of what HAMP is.  As far as I know, it all started all with the wish from the group around Gerhard Peters in Hamburg to mount a Ka-band cloud radar on HALO.  As a belly pod had to be constructed anyway, it was decided to make it actually larger than just 1x1 m for the radar antenna in order to also accommodate passive microwave instruments.  Finally, three boxes containing five radiometer receivers were designed and accompany the radar to form the HAMP: the HALO Microwave Package.

Regardless of the troubles with one of the three radiometer boxes, already a couple channels of the two other boxes are enough to keep Heike and me busy and to provide information, that other instruments can't.  For example, when specMACs is blinded by the white beautifulness of a cloud, HAMP can come along and tell you how watery your flowers are. The seven K-band channels and the 90 GHz channels can be used to retrieve the LWP of the liquid water in and below the cloud.  The combination with the radar enabled us to separate the total LWP by its contributions due to small cloud droplets and bigger rain droplets. Hence we also estimate the rain water path (RWP).

Since NARVAL2, we have been working on quality assessment, bias corrections, retrieval development and accuracy assessment to provide continuous insights on integrated water vapor, liquid water path and the separation of cloud and rain water.  I will present parts of that on a poster during UCP2019. So come along to discuss our progress (poster 197)!

As Tobi pointed out, the spatial 2D view from specMACS plus the time provides somehow a three dimensional impression of the clouds below HALO.  Thought as a multi wavelengths flight entertainment system, I build the video graph a while ago.  Video graph takes specMACS videos, as also shown in Bernhards mail, and displays other measurements alongside.  So the visual impression can be accompanied with microwave observations.  As HALO is flying, clouds pass the specMACS field of view from bottom to top and so do the plots next to the video in video graph.  I tried do align the time series of the nadir sampling HAMP with the clouds in the camera field of view that transected the nadir beam. Corresponding to the cloud movement, y is the time axis.  In todays case, LWP and RWP time series on top of a hight-time plot of the radar reflectivity is shown as data.  The x-axis indicates the LWP/RWP magnitude.  The radar image has its bottom (sea surface) on the left and top (flight altitude) on the right hand side of the plotting area.  This is also explained in the attached screen shot.

As the entry point for todays entertainment, I took one of Bernhards glory sections (15:43:26):


In this case, the glory clouds are thin enough to not overexposure specMACS and thick enough to be seen by the radar, as well as containing liquid to overcome the 10 to 15 g/m² LWP sensitivity threshold of the radiometers.  This video animation shows how visible and non-human-visible radiation complement each other.  Later you see blinding cloud beauties and with their vertical structure in the radar and water content as LWP.

I am curiously looking forward how the clouds will look like during EUREC4A, especially when we finally have an infrared imager telling us more about the cloud top temperatures and thus their buoyancy and maturity.  In the meantime, also the HAMP instruments experienced an overhaul. We got rid of the most unstable channel (outermost at 183 GHz).  The hardware of remaining channels and other radiometer modules got replaced, repaired and upgraded.  Amplifier and detectors became better and so the brightness temperatures noise is reduced.  We will try how this can provide higher spatial resolution as the sampling frequency is four times faster now.  Let's keep our finger crossed, that these upgrades will give more than just high resolution zig-zag issues ;)

All the best,

*) Works best in up to date Firefox and Chrome browsers.  Click the play button twice (start-stop) to trigger the video stream.  If the Video is running after loading but the plots stuck, try pausing and play again.

Back to top