Elements for Life

[

] 92

warnings, hydrology (including flood control), air-quality

forecasting, chemical emergency response, transportation

safety, and energy management… The Helsinki Testbed

is a

Finnish initiative aimed at developing enhanced three-dimen-

sional mesoscale observation networks critical to the

advancement of modeling systems and related user applica-

tions. It is a public-private-academic partnership. The

program is open to all interested parties and the data is freely

accessible through the Internet. Related stakeholder groups

include homeland security, agriculture, insurance, urban

management, coastal zone management, media, and public

safety.”

A recent community workshop

with international partic-

ipation considered the requirements of effective mesoscale

measurement networks, and concluded that: “existing

mesoscale measurement networks do not provide observa-

tions of the type, frequency, and density that are required to

optimize mesoscale predictions and nowcasts. To be viable,

three-dimensional mesoscale observation networks must

serve multiple applications, and the public, private, and acad-

emic sectors must all actively participate in their design and

implementation as well as in the creation and delivery of

value-added products. The [urban] measurement challenge

can best be met by an integrated approach that considers all

elements of an end-to-end solution: identifying end users and

their needs; designing an optimal mix of observations; defin-

ing the balance between static and dynamic (targeted or

adaptive) sampling strategies; ensuring data standards and

data quality, establishing long-term testbeds (such as evalu-

ation and demonstration programs); and developing effective

implementation strategies.”

The challenge is to determine the most effective mix of obser-

vations, including alternative network configurations and

sampling strategies. For example, in improving mesoscale

analyses and predictions, it may be more cost effective to

sample only the boundary layer, with denser coverage, than to

similarly enhance observations in the upper troposphere. It

may be more cost effective to deploy intermittent, targeted

observations at high resolution than to maintain dense arrays

of continuous sensors. Regional testbeds are an intermediate

step needed to provide answers to these and other questions.

Testbeds must carefully gauge the value of forecast products

provided to end users.

Improved mesoscale observations present many chal-

lenges. For example, the top observational priority for

operational nowcasting is to establish a dense mesoscale

network of surface weather stations to measure winds and

state variables and provide real-time sub-hourly reports.

Minimum station spacing in urban areas should be 10 km

or less, and the reporting frequency should be every five

minutes or less. Radar is an invaluable tool for nowcasting

applications, yet the current operational systems have not

kept pace with technological advances. Dual-polarization

capability should be implemented on existing radars, and

private and academic radars should be integrated into oper-

ational networks. Consideration should also be given to

deploying X-band polarimetric radars, as well as techniques

for improving boundary layer coverage through the use of

closely spaced, low power X-band radars. Radar refractivity

measurements should be evaluated as a possible tool for

improving nowcasting by sensing moisture discontinuities.

Products detailing near-surface water vapour fields should

be provided in real time to forecasters and assimilated into

models to demonstrate their potential to improve nowcast-

ing. There is also a pressing need to provide boundary layer

observations using radio frequency (RF) wind profilers. Not

only are additional observation systems required – includ-

ing in situ and remote sensors, both earth- and satellite-based

500

1000

1500

2000

2500

3000

3500

4000

4500

1950

Population (millions)

Urban-more developed regions

Rural-more developed regions

Urban-less developed regions

Rural-less developed regions

1955 1960 1965 1970 1975 1980 1985 1990 2000 2005 2010 2015 2020 2025 2030

Urban and rural population of the more and less developed regions, 1950-2030

Source: UNPD, 2001