In 2008 the Red Cross Red Crescent (RCRC) started with Forecast-based Financing pilots to improve existing Early-Warning Early Action systems. Forecast-based financing is a new methodology to prepare, deliver and respond in a more effective and efficient manner, based on hazard forecasts. Actions are triggered when a forecast exceeds a danger level in a vulnerable intervention area. Forecast-based financing consists of several implementation steps, of which the first three aim at impact-based forecasting. Therefore, In this study we investigate how forecast skill of agricultural drought forecasts can be achieved. More specifically, the aim is to identify the contribution of machine learning and satellite-derived products in early warning early action systems improving the forecast skill of agricultural drought forecasts. We explore this through a machine learning model for a case-study area of the Lower Shire River Basin in Malawi. Several experiments with different sets of predictors and predictands are conducted to test which data adds to the skill and at what spatial detail. As predictors, the following agro-climatic indices are used: cumulative precipitation, soil moisture anomalies,mland surface temperature anomalies, El Niño Southern Oscillation in July and four different dry spell categories within the growing season (0-2 days dry spell, 3-4 day dry spell, 5-10 day dry spell and larger than 10 day dry spell). As drought predictand, the normalized difference vegetation index (NDVI) and the vegetation optical depth
(VOD) in March are used, the latter obtained from satellite data company VanderSat. The final set of predictors and predictands is narrowed down based on which data is available and with which quality (timeliness, reliability, accuracy). Initial results, show higher accuracy and weighted accuracy values for the models including soil moisture data compared to the ones without soil moisture, expect for the last month in the growing season, where it give opposite results. The outcome of the model can support humanitarian organisations to increase the lead time necessary to act upon a drought trigger and reduce the impact of such event.

The Philippines is a country located in the typhoon belt of the Pacific. Its location makes it typhoon prone and on average around 20 typhoons enter the country every year (Asian Disaster Reduction Center, 2008). This natural phenomenon can have disastrous effects and therefore, good preparedness is of special importance. Traditionally weather forecasts have been used in order to predict the physical characteristics of typhoons to organise early actions that can dampen their damage. However, regardless the good meteorological forecasts, typhoons kept having large effects in coastal areas, due to the gap of knowledge between a hazard and its impact. For this reason, nowadays disaster managers are increasingly more interested in the knowing the repercussion of natural hazards, which can help to prepare and mitigate their consequences (Tozier de la Poterie et al., 2018).
Although forecasts that predict the impact of hazards, the so-called impact-based forecasts, are rising, there are not many of these systems operationalised yet. To make these forecasts more functional it is essential to understand what their users’ needs are. Because disaster managers work under stressful conditions, it is crucial to have the right information in order to take effective actions. This research has developed an impact-based forecasting prototype for Manila Bay, Philippines, considering the information needs of disaster managers. In order to do so, an iterative process has been followed in the creation of the system, in which disaster managers provided their inputs and feedbacks on the prototype. The impacts on the assets of relevance for the disaster managers have been forecasted with the use of Delft-FEWS and Delft-FIAT. Furthermore, this forecasting system has been novel in predicting the impact of typhoons considering not only the hazard and exposure but also implementing the vulnerabilities of the affected area. The findings of this study suggest that an impact-based forecasting system for Manila Bay should provide information on the affected population, livelihoods, hospitals, roads and schools. Those assets have been pointed out as most important for disaster managers at local scale. The output of the forecasting system should provide actionable results that allow disaster managers to make quick and relevant decisions. Furthermore, the data displayed in the system should be simple, clear and with colour codes that allow for a fast interpretation of the results and provide maps with information of the impacts on all the assets per municipality or province. Besides, it has been observed that vulnerability plays an important role when prioritising the action areas of disaster managers. The aim of this research was to develop an impact-based forecasting system that considers the information and display needs of disaster managers, while integrating the vulnerabilities of the affected area. Therefore, the objective of this thesis was not to develop an accurate mathematical model that provides the exact impact. Instead, it is a guide of what a disaster managers-based forecast should provide. Hence, in order to make this forecasting system operational, validation of the prototype and consideration of other hazards, such as wind speeds, are recommended.