Good to see some in-depth and constructive comments. Has anything else come in since 16 June?

James

Bruce

Page 6: For the purpose of flood forecasting and warning I would add the requirement of humidity and wind speed. Firstly most methods used in practice utilise soil moisture accounting models – which require evaporatranspiration as an input. Also for catchments influenced significantly by snow, these variables can be important for describing snowmelt.

Page 8: The number of days after which no skill can be derived for decision making seems very crisp. This is very much dependent on the size of the basin as well as the dominant climatology. In small basins (e.g. 250,000 km2) there may be quite some skill in these longer range predictions. Also dominant processes such as snowmelt may be of significant influence on the predictability of available water resources at longer lead times. I would in any case add some note of this relationship with size of basin, as well as dominant processes.

Page 12: In the last line of the 1st paragraph and also in the second paragraph, the argument seems to be given that data should not be shared between agencies – and if so only at a cost and under contract. It also suggests that duplication is not necessarily an issue, in lieu of independence of agencies. I am not sure I would fully support this argument – and would suggest promoting a more open approach – such as the paradigm followed in the US. I believe that a lot can be achieved by sharing data more openly, in particular in developing countries where data ownership is seen as a proxy for power, and in many such countries payment schemes such as also suggested later in the document lead only to virtual inaccessibility to data.

Page 29: In the sections on data sharing, rationalisation of networks and modernisation, I would emphasise the importance of seeking to use open standards for data – as well as for meta-data. The use of standards (some currently emerging) will greatly enable interchange that has been hitherto difficult. There have been some recent efforts to improve the use of standards in the water resources community – but it would seem that there is much to learn from the meteorological community in this respect. Standards in meta-data are equally important as without this the sharing of the data is often meaningless.

In the last section on issues and applications, it may be relevant to come back to some of the issues noted in the 2nd section, in particular the difficulties of measuring the primary variables that are used in water resources management. Whilst some of the issues with remotely sensed data are very rightly mentioned, the opportunity of improving the ground based data is not. In particular in mountainous basins and even more so in mountainous basins with significant snowfall, the characterisation of precipitation is extremely difficult. Point rainfall measurements may be underestimate precipitation by 30-40% in such regions, and I think there are very few cases where the density of the network is adequate to capture the spatial variability of precipitation. Observation of flow (and connected to this management of rating curves) is another key issue, and equally very often neglected.

Table 3.1 – Flood plain zoning/flood frequency estimation – We do understand that flood frequency analysis is the need for flood plain zoning. From that perspective the observed discharge data are used in most of the cases. If an outlook for future changes is made than we think that beside precipitation and PET an air temperature must be included (to cover snow melt dominated streams). In addition synoptic information (frequency of synoptic patterns) might be of good use.
Correspondingly for navigation air temperature should be included to address a snow melting.
We also propose to ad medium and long term forecasts to flood management and control especially for spring melting floods and some regions with high dependency of large scale patterns such as ENSO.

Table 3.2 – From our perspective we can not clearly distinguish between strategic and tactical categories used in table (and listing particular feature under those categories).
We propose to delete (extreme precipitation, temperature) from tactical-climate (6-12 m) cell. Extremes are not covered by seasonal forecasting and thus proposed wording could be misleading.
We also propose to move risk mapping from climate column to long term change. We are not aware of risk re-assessment in 6-12 month scale being done in practice. However it might be a case in regions with strong effect of large scale phenomena such as ENSO (if this is a case it should be explained in more detail).

p.11 – Please check following sentence for clarity: “Evaporation data are produced by measurement by evaporation pans or evaporimeters, or estimated a evapotranspiration.”

Box 1. – We propose to enlarge the examples by including another two.
1 – there is no example of the best solution – a common hydrometeorological service under one institution (like NWS in the USA) or preferably in one department or office (like Swedish Meteorological and Hydrological Institute – SMHI or Czech Hydrometeorological Institute – CHMI). Such arrangement enables to develop common formats and share the products and tools developed to reflect both points of view meteo and hydro. Users of the meteorological warnings and flood warnings are overlapping (civil protection, general public) and one set up of warning system (format, links etc.) is of great advantage (i.e. Meteoalarm inclusion of flood warnings is prepared easily).
In addition that is the easiest way to bridge the gap between the background and scientific language.
2- for those countries not having joined NMHS a French solution (establishing the SCHAPI office) could be interesting example.

p. 30 – The last bullet definitely could not be reasonable agreed. No meteorological service is able to guarantee its single (nor monthly or annual) forecast performance!!!!. In most cases, they will agree to attach the information on long term performance to illustrate the average uncertainty.
On the other hand a data policy applied to the user should be mentioned including the issue of interpretation of forecasts.

p. 32 – Not all instruments are covered by WMO standards so far (ADCP could be an example – standards are under the development).

Chapter 5.3 – We propose to mention there the need for calibration especially for electronic equipment, as sensors may lost the declared accuracy in time, because of the construction and sensitive components being included.

Chapter 5.4 – We propose to mention a option of a web services (user ask for the data actively). Such system was implemented by CHMI for providing hydrological data from automatic water gauges (all data are freely available at web page: http://hydro.chmi.cz). Defined users may download a xml file with data. Using web service enables to skip agreements and operation control by the donor.
We also propose to mention largely used database system CLIDATA (http://www.clidata.cz).

Chapter 5.5 – We do not agree with last sentence. Our experience proves that web is becoming more and more important in distributing the information. In designing the web application and hardware structure the action could be done to decrease the black outs to reasonable amount (comparable to other electronic ways of data distribution).

p. 34 – Last paragraph should be titled as summary, or conclusion as it is relevant for the whole report not only chapter 5.7.

Key elements of our quality management include the monitoring of uncertainties within the data collection, adherence to standards and the efforts required to improve technical processes and information with developments in technologies. Staff development is vital, and quality management tools are helpful in this area.

We see quality management to be very much an on-going process, and would look to integrate elements of WMO’s QM Framework as applicable.

Within the ISO 9001 there is a need for having guidances on all processes within the service, those guidances on hydrological proceses are mostly internal documents. Therefore CHMI would use QMS hydrology to implement it in to those documents to respect the WMO standards defined by QMS.