Climate Exchange

[

] 159

isaster

isk

eduction

develop an insurance against water deficit and achieve reinsurance

from an international insurance company.

The development of the insurance involved a second step: the

calibration and validation of the parameters of the hydrological

model over the period 1981-2010. Here, historic daily rainfall data

and average monthly evapo-transpiration of the basins was used.

DNM, INMET and the Joint Technical Commission of Salto

Grande (URU-ARG) provide daily fields of precipitation over the

basins of the Uruguay and Negro rivers. Homogenization of the

data had to occur across the borders, as data from two different

meteorological services were used.

UTE provides the analysis and calculations for rainfall-flow trans-

formation on these basins in order to estimate using a hydrological

model and the expected inflow to Salto Grande (Uruguay-Argentina),

Gabriel Terra and Constitución (Uruguay) hydroelectric reservoirs.

The hydrological model proposed is the Temez model. This is

a model of four parameters, expressed in an aggregated format

over the Thiessen polygons of each rain gauge, since it works with

average values of variables and parameters and reproduced in a

simple and conceptual hydrological cycle in monthly steps. The

model considers two layers, essentially representing the upper

unsaturated zone and the lower saturated zone. These two layers

can be considered as storages, generating fast and slow surface

flows respectively.

Monthly precipitation in a rain gauge used in the Temez model

is time–displaced, to take into account the delay of the flow from

the rain gauge out of the basin. The total inflow of a basin is then

calculated as a sum of the flow derived of the Thiessen polygons

of each rain gauge in the basin. With the estimated total inflow in

each basin in a semester, the UPHEI can be calculated.

The insurance was defined in terms of the values

of UPHEI in the semester. If the estimated UPHEI is

smaller than a defined trigger, a corresponding payment

will be made; otherwise the payment is null.

In a third stage, during the operational phase, the

insurance company requires semi-annual forecasts of

monthly discharges in the basins. Daily rainfall records

of selected stations, at the end of each month during the

operation, will be used as input for the hydro-electrical

model. DNM has established a standardized rainfall

index (SPI) which will define the occurrence of drought

events. The SPI is an index based on the probability

of recording a given amount of precipitation, and the

probabilities are standardized so that an index of zero

indicates the median precipitation amount (half of the

historical precipitation amounts are below the median

and half are above it). The index is negative for drought,

and positive for wet conditions. As the dry or wet condi-

tions become more severe, the index becomes more

negative or positive. The SPI is computed for several

time scales ranging from one month to 12 months,

to capture the various scales of both short-term and

long-term drought. Inferences of shortcomings in the

upcoming six months can then be made in advance and

preparatory measures can be taken. At the same time the

insurance company can prepare for potential payments.

The insurance will work in a pilot phase until 2016.

An extension will then be considered depending on the

successful implementation of the system.

Ongoing provision

The provision of this service requires the combina-

tion of the capabilities of at least four institutions:

DNM, INMET, URU-ARG and UTE. For the effec-

tive and timely delivery of this service by DNM, the

hydrometeorological database (in operation since

2010) was widely used. A team of external engineers

and programmers supported this database. As well as

developing an effective and easy-to-use maintenance

system, they also expanded the data extraction, trans-

formation and processing routines of the available

rainfall information. Financing for this team came

from state resources transferred to the Dirección

Nacional de Meteorología.

Future goals for the service include improving the

availability of rainfall data in the basins under study,

by incorporating automatic stations and satellite rain-

fall estimation. New automatic pluviometer stations

have been acquired in the Negro River basin by UTE,

and URU-ARG has also acquired new rainfall stations

for the Rio Cuareim basin. DNM plans to address the

installation of automatic stations in cities such as Salto,

Rivera and Bella Union.

In future, the service could be expanded to include

other countries and other sub-basins of the Rio de la

Plata basin. An extensive cost-benefit analysis is being

conducted by the insurance company which will

provide insight on the economic value of providing

this climate service.

Source: UTE (Uruguay)

The Temez conceptual model: A surplus (T) part of the water which

precipitates (P), is infiltrated (I) to the second layer, and goes to the

stream-flow (Asup), whiles the rest of this water (P) is stored in the first

layer of soil. Here, evapotranspiration occurs and the soil may also store

water from one month to another. In the second layer, the water can

drain to the stream-flow and may also be stored

The Temez conceptual model