Climate change modelling and analysis for Mozambique
This article summarises key points from a report by Mark Tadross, of the Climate Systems Analysis Group at the University of Cape Town, titled Climate Change modelling and analysis for Mozambique (March 2009). The study and report was commissioned by Mozambique’s National Disaster Management Institute, the Instituto Nacional de Gestão de Calamidades (INGC), for their adaptation to climate change project.
Study
The climate messages presented here are based on analysing: observed data collected from 32 weather stations across Mozambique since 1960; the outputs from 7 global circulation models (GCM) i.e. projections of future climate conditions; and the results of downscaling these GCM outputs to a number of station locations (using the observed data record to do this) for 2 periods in the future: 2046-2065 and 2080-2100. The climate impact messages are developed out of the results from various impact models (flood and crop-suitability), run using the observed climate data and the downscaled projected climate data (i.e. scenarios of future climate over Mozambique for middle and end of this century).
Please note that this work only covers the biophysical aspects related to climate and climate impacts and does not touch upon the social and institutional factors that result in differential vulnerability and thereby the risk these climate conditions and climate impacts pose to different groups of people.
Characterising Mozambique’s climate
- predominantly maritime climate
- climate largely determined by offshore warm waters of Agulhas current and tropical cyclones which typically pass from the north to the south
- generally temperature are warmer at the coast and cooler inland as high altitudes
Intra-annual variability (within a year):
- important weather systems that determine rainfall patterns over Mozambique are: the Inter-tropical Convergence Zone; tropical cyclones; thermal lows; and incoming easterly waves
- seasons: rainy and hot season Oct – April; dry and cooler season May – September
- rainfall also influenced by local variations in altitude
- heaviest rainfall associated with passage of tropical cyclones (active Nov – April, especially Jan/Feb) – often resulting in severe flooding
Inter-annual variability (between years):
- associated with El-Niño Southern Oscillation (El-Niño years associated with drier conditions)
- influenced by sea surface temperatures in the Indian Ocean
Key climate messages
Historical Temperature:
- the observational record shows a trend of increasing temperatures over most of the country in all 4 seasons
- these trends are particularly marked in minimum temperatures i.e. temperatures are not dropping as low now as they did earlier in the 20th century, for example in the central regions of Mozambique the minimum temperatures experienced in winter have increased by 1.62°C between 1960 and 2005
- this in turn means that the number of cold nights and cold days have decreased over this period, while the number of hot mights and days has increased
Historical Rainfall:
- clear trends in rainfall are more difficult to detect from the records
- it looks at though there is a later start to the rainy season e.g. in certain locations in the north of the country the start of the rainy season has been delayed by up to 45 days comparing years for the period 1960 to 2005
- it also looks like there is an increase in the number of consecutive days without rain in the north-east of the country during the shoulder seasons i.e. March – May and September – November
Future Temperature:
- all the models show an expected increase in temperatures between now and the middle of the century (i.e. 2046-2065)
- these increases in temperature are expected to be higher inland than in the coastal areas
- it seems that temperatures in the September – November period i.e. spring will increase by the most – an increase of 2.5 – 3°C in some places
- by the end of the century (i.e. 2081-2100) spring time temperatures in the central region may have risen by as much as 5 – 6°C
- the models suggest that the likelihood of experiencing extremely hot days (that is days with a maximum temperature over 35°C) will increase over all parts of the country, by about 7% by the middle of the century and 25-33% by the end of the century
Future Rainfall:
- downscaled projects from 7 GCMs suggest an increase in rainfall in the December-May period between now and the middle and end of this century, particularly in the coastal areas, BUT the spread of models is large, in other words some are showing a big increase, some a much smaller increase, and a few even suggesting a possible decrease
- increases in rainfall are likely to be greatest towards the end of the summer season, especially in the north and central regions
Linking Temperature and Rainfall:
- in order to try and understand the possible combined effect of changing temperatures and rainfall in a given area researchers explore changes in evaporation rates i.e. how quickly water is ‘lost’ from liquid on/in the ground (that humans, plants, animals easily use) to vapour in the air
- for all regions in Mozambique it seems that increases in evaporation will be greater than increases in rainfall during the winter and early summer (June-November), which means that the dry season will be even drier (especially over the central region)
Important challenges and considerations related to these climate messages
- limited coverage of station data over the country (especially in Gaza and Tete provinces), with an average station density of 1 station every 29,000km2 (as compared with 1 station every 1000km2 in South Africa) although there is obviously not an even spread (the majority are along the coast of Mozambique) – this data is used for downscaling projections of future climate, only 27 of the 32 stations had sufficient temperature and rainfall data for statistical downscaling to be applied
- stations where there are substantial amounts of missing data in the time series can not have trend analyses done
- data quality is poor so much time has to be spent cleaning the data before any analysis can be done
Key messages on climate impacts
Important challenges and considerations related to these climate impact messages
- using a model to try and determine the impact of these climate changes on crop suitability, and thereby agricultural production, is made difficult by needing the correct baseline data to add the climate projections to, and the assumptions that have to made in calculating potential evapotranspiration
- the modelling of associated flood risks is made difficult by the limited coverage of the climate data from monitoring stations (this is dealt with by using satellite-derived gridded data) and by the the challenges of simulating maximum DAILY values for the future (i.e. the time resolution relevant to flooding events)