| 03 Mar 2026
Enhancing GNSS Water Vapor Retrieval via Synergistic Microwave Radiometry: Thermodynamic Error Diagnosis and Bias Correction
Abstract. The retrieval of Precipitable Water Vapor (PWV) from Global Navigation Satellite Systems (GNSS) in thermodynamically complex environments is fundamentally limited by the accuracy of the weighted mean temperature (Tm). This study evaluates the efficacy of static climatological models versus dynamic ground-based microwave radiometry for Tm determination in the Eastern Mediterranean, a region characterized by sharp refractivity gradients. Using the Cyprus GNSS Meteorology Enhancement (CYGMEN) infrastructure in Nicosia, the performance of the ERA5-based HGPT2 model and a co-located Microwave Radiometer (MWR) was assessed against radiosonde (RS) profiles during the 2025 warm season (Spring–Summer). Diagnostic analysis reveals that the static HGPT2 model fails to resolve the diurnal thermodynamic decoupling between the boundary layer and the free troposphere, leading to a systematic overestimation of Tm exceeding 6 K during peak solar insolation. Conversely, the MWR captures short-term thermodynamic variability (r=0.98) but exhibits a systematic cold bias of −1.91 K in raw retrievals. It is demonstrated that a site-specific linear bias correction reduces the MWR Tm Root Mean Square Error (RMSE) from 2.32 K to 1.43 K, significantly outperforming the empirical model. Sensitivity analysis confirms that thermodynamic uncertainty dominates the error budget, outweighing uncertainties in refractivity constants by an order of magnitude. Consequently, standard climatological retrievals diverge from the synergistic MWR-GNSS method during extreme hygrometric events, introducing systematic PWV biases exceeding 1.0 mm when moisture levels surpass 45 mm. The synergistic coupling of real-time radiometric Tm with GNSS data is therefore essential for generating climate-quality PWV records in semi-arid coastal regions.
The work presented in the manuscript address the possibility to improve the accuracy of the conversion from Zenith Wet Daley (ZWD), estimated from observations with a ground-based GNSS station, to Integrated Water Vapour (IWV) by adding data from a microwave radiometer capable of providing information about the temperature profile in the atmosphere.
Although these results are of course unique for this specific station, The findings are primarily of interest for users that have (or are considering tu buy) this type of radiometer. It is a much larger investment compared to the GNSS ground station itself. In any case, because as far as I know it is a new concept, and it makes sense that the study is made available for the entire scientific community.
My criticism is that when reading the manuscript it gives the impression that it is the uncertainties in the mean temperature that is the limiting factor for the quality of the IWV. I like to point out that even during extreme conditions (regarding the temperature profile) the uncertainty in the final estimate of the IWV, the Zenith Total Delay (ZTD) is at least as important. See e. g. Table 4 in:
Ning et al. (2016). The uncertainty of the atmospheric integrated water vapour estimated from GNSS observations, Atmos. Meas. Tech., 9, pp. 79-92, https://doi.org/10.5194/amt-9-79-2016.
Below I suggest alternative wordings in order to nuance this issue.
Rain is only mentioned in terms of extreme weather in the introduction, but it is not pointed out that the microwave radiometer algorithms more or less break down during rain. I assume that the station on Cyprus may not be exposed to rain during a large percentage of time, and perhaps the extreme temperature profiles (difficult to model based on ground temperature only) which the study is focused on, do never occur during rain? In any case, I think it is important to mention the poor accuracy of the radiometer retrievals during rain.
Specific comments
Line (L) 10: fundamentally limited --> significantly affected
L 23: essential --> meaningful
L 57: Here you state that "Tm is the primary source of uncertainty in GNSS meteorology after ZTD estimation". It will be appropriate to either (1) also here point out that the uncertainty in the ZTD is larger or comparable when it comes to the impact on the uncertainty of the final IWV, or (2) replace "the primary" with "one".
L 60: Assuming that 1-2 % error corresponds to 1-2 mm in the PWV implies that the value of the PWV is 100 mm. I am not sure if such a high value has ever been observed. Why not just state the percentage uncertainty and let the reader figure out what it menas in absolute value given the whether conditions.
L 84: It will be interesting to know how many sites with co-located MWR and GNSS there are in the network?
Table 1: It will make sense to also state that the MWR "Role in Study" is also to provide the IWV.
Table 1: You state that the humidity uncertainty of the RS humidity sensor is 4 %. This deserves some comment in the text when you use this sensor as "ground truth" given that you evaluate uncertainties at comparable and lower levels.
L 155: The IGS Ultra-Rapid products are not the optimal choice when aiming at the lowest possible uncertainty which should be the case for climate studies. Please comment on the impact of this choice and why it was done this way.
L 182: What are the estimated values of alpha and beta?
Figure 4(b): Which estimate of Tm is used in this comparison?
L 299: This result is already stated in L 288.
L 341: Here again it is implicitly assumed that the error in the ZTD is zero, which is not true. Please rewrite.
Figure 9: This figure does not give the whole picture. Even if you do not include additional errors in the figure, I think it is necessary to mention them. I am thinking of orbit and clock errors, as well as mapping function, and site dependent (the electromagnetic environment such as reflections/multipath) errors in the GNSS processing.
L 475: Give a reference for the "known artifact".
L 510 - 514: "This correction effectively halves the uncertainty in the final GNSS water vapor product compared to standard climatological approaches." is not true, and it is not "the dominant error source in GNSS meteorology". These statements ignore the errors in the ZTD from the GNSS processing. When rewriting this text I suggest that you instead use words such as "may be comparable to", "are significant", or something similar.
In this context I think you also have to point out that there are many sites world-wide where the use of microwave radiometry may not be motivated (given the high cost, all the other uncertainties, and that models for Tm is not that poor everywhere).
Technical Corrections
L 27 + more: You use the American spelling of vapour, although ACP is a European journal?
L 30: Here you introduce PWV (unit mm). Later (L 118) you also use IWV (unit kg/m2). You should use only one of these in order not to confuse the reader. I suggest to chose IWV because the delays from GNSS is expressed in unit of length and it makes it easier to see the difference.
L 34: Here the acronym EM is defined, but it is defined again on L 464. At many other places you do not use the acronym. The acronym is only used 9 times, so I suggest not to use it at all. I think it will make the reading better.
L 60 + more: 2% --> 2 % (SI recommendation)
Table 1: 1 sec --> 1 s (SI recommendation)
L 114: Training Set --> training set
L 115: Validation Set --> validation set
L 132: delete "total column" (it is already in the definition of the IWV)
L242: 6.7°C --> 6.7 °C
L 284: whereas the MWR exhibits --> whereas it relative to the MWR exhibits
L 560, 567: doi missing
L 589: the doi link is not correct, shall be:
https://link.springer.com/book/10.1007/978-3-030-13901-8, and the author list is not correct.
(Note that I have not checked all the doi links, so it may be a good idea for you to carry out?)
L 621: add https://doi.org/10.1175/1520-0442(1996)009%3C3561:TWVCAT%3E2.0.CO;2