Acknowledging a limited knowledge and use of climate data in development economics, this thesis proposes a new perspective on the use of both macroeconomic and microeconomic climate data. This thesis has a two-fold objective: to deepen knowledge on climate data through the production of a new global database directly exploitable by economists, and to propose several applications of this database at different economic scales. The first part focuses on the macroeconomic aspect of the climate data. Its aim is to empower researchers with a deeper understanding of global climate data and to increase economic knowledge on the characteristics of climate change. Chapter 1 contains a climate database available monthly over the 1900-2008 period for nearly 200 countries. Based on this initial work, Chapter 2 proposes an original use of the database with the construction of a physical indicator of vulnerability to climate change, a tool which can be used as a basis for the allocation of aid for climate change adaptation. The second part of this thesis is composed of two microeconomic studies whose aim is to analyse the behaviour of agents facing a rainfall shock. In those two chapters, rainfall data is used to compensate for the lack of data on agricultural production. At the household level, Chapter 3 focuses on the implications of an episode of reduced rainfall on the nutritional status of children under five years old. It shows that rural households do not have the capacity to provide for or absorb climate shocks. At the grain market level, Chapter 4 studies the response of prices to a shock front rainfall. It reveals the presence of speculation, especially in small and poorly integrated markets.
