DPFS_Manual_Appendix-II-2

OBSERVATIONAL DATA REQUIREMENTS FOR GDPFS CENTRES FOR GLOBAL AND REGIONAL EXCHANGE

The following paragraphs 1, 2 and 3 state the observations required to operate all GDPFS centres at national, regional and global levels. Paragraph 4 addresses the data requirements for NWP operations only.

1. The types of observation networks and platforms providing data required at data-processing and forecasting centres are as follows:

(a) All stations included in the Regional Basic Synoptic Networks;

(b) The network of supplementary synoptic stations, including automatic stations;

(d) Mobile sea stations;

(e) All other stations making radiowind, radiosonde/radiowind and pilot balloon observations;

(f) Meteorological rocket stations;

(g) Aircraft meteorological observations;

(h) Wind profilers;

(i) Doppler and weather watch radar systems and networks;

(j) Space-based systems producing:

(i) Imagery (including both analogue and digital imagery);

(ii) Radiance data;

(iii) Vertical temperature and humidity profiles;

(iv) Cloud and water vapor motion winds;

(v) Cloud height, temperature, type and amount;

(vi) Digital information about clouds (liquid water or ice (total));

(vii) Surface wind, precipitation rate and precipitable water;

(viii) Land temperature;

(ix) Sea-surface temperature;

(x) Ocean surface wind vector;

(xi) Albedo;

(xii) Ocean wave spectra;

(xiii) Sea ice cover;

(xiv) Snow cover, depth and water equivalent;

(xv) Earth radiation fluxes;

(xvi) Aerosols and trace gases;

(xvii) Volcanic ash;

(xviii) Other meteorological and environmental products;

(k) Radiological data reporting stations in case of nuclear accidents (required for GDPFS centres running transport models for environmental emergency response);

(l) Selected climatological/agrometeorological/hydrological stations;

(m) Lightning detection and location systems network;

(n) Global Atmosphere Watch (GAW) network.

The observational data which will be needed to obtain optimum results from NWP systems by the year 2000 and meet the needs of all WMO Programmes and WMO supported Programmes are elaborated in paragraph 4 below and its related three tables.

2. The report code types which carry the data provided by the platforms listed in paragraph 1 above are given below:

(a) BUFR and GRIB;

(b) TEMP — Parts A, B, C and D;

(d) TEMP SHIP — Parts A, B, C and D;

(e) PILOT SHIP — Parts A, B, C and D;

(f) TEMP MOBIL — Parts A, B, C and D;

(g) PILOT MOBIL — Parts A, B, C and D;

(h) COLBA;

(i) TEMP DROP;

(j) ROCOB;

(k) SYNOP;

(l) SHIP;

(m) Reports from automatic stations on land and at sea;

(n) CODAR/AIREP/AMDAR;

(o) Selected satellite data, such as cloud images, SATEM, SAREP, SARAD, SATOB;

(p) BUOY;

(q) CLIMAT, CLIMAT SHIP;

(r) CLIMAT TEMP, CLIMAT TEMP SHIP;

(s) BATHY, TESAC, TRACKOB;

(t) WAVEOB;

(u) RADOB;

(v) RADREP.

NOTES:

(1) Items (a) to (v) do not indicate priorities.

(2) BUFR and CREX can encode any of the other above data forms and many more. If BUFR or CREX is used to represent any of these data forms, in lieu of the specific alphanumeric code form, the same data requirements apply.

3. The frequency of observational reports required is as follows:

(a) BUFR and GRIB, as available;

(b) TEMP, PILOT, TEMP SHIP, PILOT SHIP, TEMP MOBIL, PILOT MOBIL, ROCOB, COLBA and TEMP DROP, as available;

(c) SYNOP, SHIP and reports from automatic stations on land and at sea — 0000, 0300, 0600, 0900, 1200, 1500, 1800, 2100 UTC and hourly whenever possible;

(d) CODAR/AIREP/AMDAR reports, as available;

(e) Selected satellite data, such as cloud images, SATEM, SAREP, SARAD and SATOB and digital cloud data, as available;

(f) BUOY, as available;

(g) CLIMAT, CLIMAT SHIP, CLIMAT TEMP and CLIMAT TEMP SHIP — once per month;

(h) BATHY, TESAC, TRACKOB and WAVEOB, as available;

(i) RADOB and RADREP, as available.

4. Data needed for advanced NWP by the year 2000 is as follows:

GENERAL CONSIDERATIONS

The following tables list the observational data which will be needed for advanced NWP systems by the year 2000. They include the needs for data assimilation and for analysis and model validation for global short- and medium-range forecasting (excluding long-range forecasting).

Requirements for regional modelling have also been considered. They have been mentioned in the explanatory text, where appropriate, but they have not been listed in the tables. Mesoscale modelling has not been considered.

It is most likely that data of the given specifications would benefit global NWP, if available; however, it does not mean that NWP could not be carried out without such data, as NWP models produce useful products even with the observational data set currently available. It does not mean either that data of higher specification would not be useful; on the contrary, when and where such data are produced they should be made available.

The problem of the feasibility of observing all the variables listed in these tables is not addressed. Most of the requirements stated here could only be met by satellite-borne observing systems. However, in many cases a combination of satellite and in situ data is needed to obtain adequate resolution and to ensure stability of calibration of remote sensing systems.

CONTENTS OF THE TABLES

The following notes provide some explanation of how the lists were prepared and some provisos on their use:

Variables

Following past convention, the observational requirements for data assimilation are stated in terms of geophysical variables. This is thought to be useful since, from a user’s perspective, these are the variables on which information is required. However, it is important to note that these variables are not always observed directly (satellite systems observe none of them directly, with the exception of top-of-the-atmosphere radiation and a Doppler wind lidar). Also, it is no longer true that the users need their data exclusively in the form of geophysical parameters; recent developments in data assimilation have demonstrated the potential and the benefits of using data at the engineering level (e.g. radiances, brightness temperatures).

Horizontal resolution

(a) In general (and with some oversimplification), data are useful for assimilation and validation on spatial scales which the models are attempting to represent. One hundred kilometres are given as the requirement for the variables listed in the tables. However, it is possible to benefit from higher resolution data, considering the current developments towards global models with a grid length of less than 50 km;

(b) Regional models attempt to represent spatial scales above the mesoscale. Observational data are required at a resolution of 10 km.

Vertical resolution

(a) The same rationale is applied here: global NWP models are expected to have a resolution of less than one kilometre throughout the troposphere and lower stratosphere, with considerably higher resolution in the planetary boundary layer. In the mid- and upper stratosphere, a resolution of two kilometres is likely to be sufficient. The requirements for observations should be comparable;

(b) For regional models, observations are required at a resolution of 100 m (50 m in the planetary boundary layer).

Temporal resolution

(a) Just as with spatial resolution, data will be useful for assimilation and validation on temporal scales, which the models are attempting to represent. In the past, this has not been the case; so-called “four-dimensional” assimilation systems would more appropriately be described as “intermittent three-dimensional” systems, and they have not been able to make proper use of observations more frequently than the period of the data assimilation cycle (typically six hours). However, continued progress towards truly four-dimensional data assimilation is making it possible to extract useful information from observations at higher temporal frequency. With such systems, higher temporal resolution can compensate, to some extent, for poor horizontal resolution when the atmosphere is moving. A requirement of three hours for upper-air data and one hour for surface data has been specified. However, like in the case of spatial resolution, upper-air data of higher specification (up to one hour) should also be made available (e.g. cloud motion wind data from geostationary satellites, wind profiles from wind profilers);

(b) For regional models, both upper-air and surface data are required at a resolution of one hour.

Accuracy

The values given are intended to represent the RMS of the observation errors. The assessment of accuracy should include not only the true instrumental error but also the representativeness error (i.e. the characteristics of some observing systems, particularly in situ systems, which sample spatial and temporal scales that are not represented by the models). For NWP applications, such effects appear as though they were observation errors.

Timeliness

In NWP, the value of data degrades with time, and it does so particularly rapidly for variables which change quickly. Operational assimilation systems are usually run with a cut-off time of about three hours for global models and one and a half hours for regional models (although data received with longer delays remain useful). Therefore, the timeliness of data delivery must take into account the advertized initiation time of any operational model that uses that data. For observations which are expected to be used for validation, and not for analysis/assimilation in near-real-time, the timeliness is less critical.

This is the static archive copy of the old wiswiki, decommissioned on June 1 2020

DPFS_Manual_Appendix-II-2