Tailoring large-scale hydrological models for national planning of climate actions in vulnerable countries

Many countries vulnerable to climate change do not yet have national modelling systems in place to guide adaptation measures. Especially low- and middle-income countries are at the mercy of global or large-scale estimates of climate change impacts, which might not be relevant to the spatial scale of societal challenges or to engineering methods based on observations.

Climate services are launched with scientific data, which can be misunderstood and misused if not communicated in a pedagogic way. For instance, the results from climate models represents an average for a calculation unit and neglects the spatial variability within that unit. In-situ observations from monitoring stations represents a point value and may thus be very different from areal estimates. Moreover, observations are relatively few leaving large areas ungauged. Sometimes, the area of interest falls in between grids or is very small compared to the grid or catchment and the average values may then not be representative or useful.

Moreover, the results from climate models represents a statistical period of 30 years, but not the chronological happening of events or weather conditions. Time-series from climate models are thus not representing specific dates and should not be compared to observed time-series but only to statistical estimates, such as indicators. 

In this presentation we showcase (1) state-of the art methods to produce climate indicators for weather and water data over large domains, and (2) some ways to tailor climate and water data for local applications and practical use.

We will demonstrate the global climate service climateinformation.org, in which climate and water indicators result from an extensive production chain, merging data from various sources with different resolution in time and space.

For water indicators, climateinformation.org uses results from a global integrated-catchment model, the world-wide HYPE. To tailor data, it is recommended to use a more detailed national/local model or set-up the HYPE model using national/local data. SMHI share the open source HYPE-model code and here we will explain how to apply climate indicators to calculate climate-change effects on water resources using a local/national model. Showcases are given for St Lucia, DR Congo, Cape Verde, and Cambodia.