Climate Exchange

[

] 61

griculture

and scientists. The specific design allows interdisciplinary working

groups to stimulate cooperation and foster knowledge exchange.

Currently, the MOSAICC toolbox is under validation in Morocco,

after which it will be implemented in other countries. The model

development is funded by the European Union programme,

Improved Global Governance for Hunger Reduction.

The system

The assemblage of the MOSAICC toolbox is based on a generic

methodology defined to assess the impact of climate change on agri-

culture, including statistical downscaling of climate data, crop yield

projections, water resources estimations and economic modelling.

Low-resolution (typically 250 km) climate data serve as primary

input for the whole model structure. These datasets feed the hydro-

logical and crop models, which need information on elevation, land

cover and soil. The outputs of these models subsequently serve as

input for the economic impact assessment.

Statistical downscaling portal for climate data

The climate scenario predictions are produced using observed

weather time series combined with global climate models (GCM).

The coarse resolution of the GCM needs to be downscaled to make

it applicable for the regional or country scale. A statistical downscal-

ing tool based on the Data Access and Downscaling tool is used.

This tool allows for downscaling large-scale predictors of data on

precipitation and minimum and maximum temperature to a set of

weather stations, provided enough observations are available.

Crop growth simulation

Two different models were selected to estimate the

crop yield. The first model is called WABAL, and was

designed to simulate the soil water balance at crop level.

Basically, it is a vertical soil profile, with no explicit

spatial extent. It is usually used at the level of agromete-

orological stations, pixels or cultivated areas. The model

requires a minimum number of inputs on precipitation,

evapotranspiration, soil water holding capacity and

crop parameters. Variables like the water requirement

satisfaction index, water surplus or deficit and actual

evapotranspiration are the outputs of WABAL. The

second model, also developed by FAO, is AQUACROP.

This model simulates the crop response to water in a

more sophisticated way. AQUACROP requires crop

parameters (data on crop physiology, cultivars, manage-

ment) that are much more sophisticated and more

difficult to acquire than the ones needed for WABAL.

A distinction is made between crop transpiration and

soil evaporation; furthermore the root development

and canopy cover can be simulated. Crop water stress,

biomass production and yield estimates are part of the

outputs. In addition, CO

concentration in the atmos-

phere can be taken into account. The selection between

WABAL and AQUACROP depends on the availability

and detail of climate information at country level.

Hydrological modelling

The Spatial Tools for River basins and Environment

and Analysis of Management options (STREAM)

is a

grid-based spatially-explicit distributed water balance

model that describes the hydrological cycle as a series

of storage compartments and flows. It was originally

designed for river basin studies with an emphasis on

water management aspects. The original version has

been successfully applied at spatial scales ranging from

small sub-basins

through medium to large basins

such as Krishna,

Zambezi

and Meuse,

up to the

(sub-) continental and global scale. STREAM calcu-

lates the water balance per time-step (user-defined,

typically 10-30 days) for a spatially distributed

gridded landscape. The model requires information

on precipitation, temperature, land cover, soil type,

depth and elevation. From temperature data the actual

and potential evapotranspiration are calculated using

the Thornthwaite equations.

MOSAICC uses the open

source version of STREAM; this version presents the

advantages of handling direct data on evaporation

and the ability to integrate dams using an automatic

calibration procedure. Furthermore, a new procedure

for flow accumulation has been implemented in this

version of STREAM.

Economic modelling

The economic modelling comprises a dynamic CGE

model that simulates the evolution of the economy

based on variations in crop yield projection.

The

model allows the user to define multiple activities

producing one commodity. To allow a spatial varia-

Source: FAO

A methodological flow-chart of MOSAICC