Evaluation of BOLAM Fine Grid Weather Forecasts with Emphasis on Hydrological Applications
, Dimitrios Koulouris 1, Ioannis L. Tsirogiannis 2 and Demetris Koutsoyiannis 3
Round 1
Reviewer 1 Report
Dear Authors
It is an intriguing study that provided valuable insights into the region of Epirus in Greece. To enhance the understanding of the area's climate characteristics and weather conditions, it would be beneficial to provide a more detailed description. Additionally, including information about the general atmospheric circulation of the region would provide a comprehensive perspective.
The analysis of 45,000 hectares using an atmospheric model is a significant undertaking. To strengthen the study, it would be important to provide a solid justification for selecting this area and size. This will help readers understand the rationale behind the chosen spatial extent.
It is crucial to establish a clear link between the soil-atmosphere interaction and the representativeness of relative humidity measurements, especially those taken at 2 meters, in comparison to the model's predictions. Exploring this relationship will enhance the understanding of how the model captures the surface-atmosphere processes and their spatial scaling effects.
Considering that this is a hydrology journal article, I understand that the topic is very interesting.
Additional comments
Figure 1. b) Please to include color scale for elevation. More description of the Figure 1.
I think it would be appropriate to include a monthly average rainfall graph (Figure 1. c).
Pag. 4, line 154. Attention: wind speed measurements at 10 meters and 3 meters provide different information due to the characteristics of the wind flow in the near-surface atmospheric boundary layer. Measurements at 10 meters may be more representative of regional or local atmospheric conditions, while measurements at 3 meters may be more influenced by the specific surface characteristics at that location.
Please to include more description for Hydrognomon software (section 2.4)
Page 17 lines 528-531 More explanation of the widespread use of Haargreaves. Pay attention to equation 10 (is it the same for daily and monthly values?)
Page 18 lines 536-544 Attention. If the Ta data in stations were compared (which fits the model well) it is obvious that the result will be good and similar. I did not understand?
When having meteorological data at the stations, it would be advisable to compare the values with the ETo method of FAO-56 (reference method). I think it would be a more consistent comparison.
Please to include lines 150-155 precision example P (mm) +/- 0.2, Ta (°C) +/- 0.1…and relevant information (operational range)
1. Precipitation (mm) (Model Rain-O-Matic, Pronamic ApS (DK) 150
2. Air temperature (°C) (Model EE08, E+E Elektronik GmbH (DE)) 151
3. Air relative humidity (%) (Model EE08, E+E Elektronik GmbH (DE)) 152
4. Solar radiation (W m–2) (Model PYR, MeterGroup Inc (USA)) 153
5. Wind speed at 3 m height (m s–1) (Model 4.3515.51.000, Thies GmbH & Co. (DE)) 154
6. Wind direction (as degrees from North) (Model NRG 200P, NRGSystems Inc (USA)) 155
Page 8 line 297-298 better explain the effect of orography
General comments
Page 1 line 13 to include Bologna Limited-Area Model (BOLAM)
Page 2 line 98 insert (Figure 1) or insert caption Figure Pag. 3 line 113.
Page 4 line 146 insert reference for WMO document. Idem for FAO.
Page 4 line 160 I cannot access of meteorological data.
Page 4 line 171 Eastern European Time (EET), Universal Time Coord…(UTC)
The measurements of each station were compared to the forecast values of the matching grid cell according to its location. This was chosen on purpose, since further manipulation of the forecast data by means of interpolation…..## I totally agree, it is very good, but attention because wind speed is different (3 m an 10 m). Please describe equation of transformation (aerodynamics profile) ## Section 2.4.
Author Response
Please find our detailed responses to the Reviewers comments in the attached pdf
Author Response File: 
Author Response.pdf
Reviewer 2 Report
The paper “Evaluation of BOLAM fine grid weather forecasts with emphasis on hydrological applications” by Malamos et al. presents a model performance evaluation on a 4 years period for some meteorological parameters. The article is well written and reports a detailed verification of a meteorological model. The methodology is applicable in general.
There are some points, however, which do not allow the publication of the paper in the current form. The first one is that the comments on precipitation are not appropriate because authors verify a hydrostatic model up to 5-6 days on a local area and for precipitation threshold greater than 20 mm/3h. It is not expected that the model is capable to predict these kind of events. It this not a matter of model error but of an error in the kind of verification required. I suggest to consider higher spatial and temporal scales for precipitation verification. Another point to take into account is the following. Considering that the model has a medium horizontal resolution, it cannot predict phenomena at local scale and post-processing technique applied to model outputs can be useful for improving prediction.
Hereafter are reported some specific comments.
- The references style is not uniform throughout the paper and sometimes it is not conform to the journal style. Please revise.
- Lines 122-123. Specify the resolution of the fine grid.
- Line 232. Based "on" the above.
- In Formula (6) and in the relative figures, what is referred as BIAS is the frequency BIAS (FBIAS). Use this name to avoid confusion with MBE.
- Lines 254-256. It is not clear the reason of using 0.1 mm as the lowest threshold to verify the model if the stations may record only values larger than 0.2 mm. I suggest to start from 0.2 mm.
- For the precipitation forecast, I think that the authors were too severe with the model. It must be considered that you are verifying a hydrostatic model at a medium horizontal resolution for precipitation forecast at local scale up to 5-6 days, which is not appropriate. In addition, you are verifying the forecast every 3 hours, and it is well known that for convective events timing errors can be larger than 3 hours. So, the precipitation forecast must be verified at larger horizontal and temporale scales compared to those shown in the paper. Stated in other terms, you are requiring to the model something that it is not able to do. Consider, for example, daily precipitation aggregated over the whole stations to have a fair comparison. Also, you are going up to 5-6 days and it is well known that the forecast has lower performances as the forecast time increases. It is very difficult that the model can catch correctly the precipitation at the local scale up to 5-6 days, which again shows that the comparison was not well done.
In addition, you showed two cases that the model missed or gave a false alarm. In a period of 4 years, this may happen several times. Again, this shows that considering the local precipitation forecast of an hydrostatic model at medium horizontal resolution for a time range of 5-6 days is not appropriate. The performance of high resolution (<3 km) forecast of non hydrostatic models can be not good for thresholds larger than 20 mm/3h. Why should you expect a good performance for BOLAM for this thresholds? Again, considering these high threshold is not appropriated for the model you are using.
Considering the conclusions about the relative humidity, they don’t seem appropriate considering the results shown in Figure 8, with the exception of day 6.
Finally, and this applies also to other parameters, consider the possibility (or at least state that in the conclusions) to improve the model forecast by post-processing.
The quality of English language is good.
Author Response
Please find our detailed responses to the Reviewers comments in the attached pdf
Author Response File: Author Response.pdf
Reviewer 3 Report
Review of “Evaluation of BOLAM fine grid weather forecasts with emphasis on hydrological applications”
Manuscript reference: hydrology-2493588
Authors: Malamos et al
Recommendation: minor revision
General Evaluation:
This work presents a comprehensive evaluation of the BOLAM weather forecast performance. In general, I find the paper well structured and well written and it presents a nice overview in the introduction, and the manuscript is a good extension of several past studies. I’m impressed by the wide inclusion of model performance metrics and the completeness of the analysis in this work. Apart from a few minor corrections suggested, I don’t find any major flaws in this manuscript and thus would recommend a minor revision.
Minor comments:
1. table 1 in the 3rd column, the title should be “number of timesteps” instead of number of data values
2. suggest use latitude and longitude instead of λ and φ in the description field.
3. table 9, 10, 11 and 12, as the units are all mm, there is no need to label each column with “mm”. to describe that in the table title is enough.
4. fonts in table 11 and 12 are different than other tables, please consider using consistent fonts throughout the manuscript.
Author Response
Please find our detailed responses to the Reviewers comments in the attached pdf
Author Response File: Author Response.pdf
Round 2
Reviewer 2 Report
The authors added some specific comments about the role of model post processing. I warmly encourage the authors to go ahead in this direction in future studies.
