Hurricanes and tropical storms occur very frequently in the Central American and Caribbean Region (CA&CR). These extreme weather events produce a lot of rain and consequently a lot of flooding. Damages and loses have been estimated to amount to 13.4 billion dollars in the last ten years. Flood Warning Systems (FEWS) are a key preventive strategy to reduce risk. Technological progress is improving the resources made available for FEWS to be more viable. In spite of the international support for FEWS and the fast development of ICT, there are very few countries in the CA&CR that have succeeded in developing fully operational warning systems that are functioning in a sustainable manner for a long period of time. There is disconnection between the community-based systems and the centralized systems, as well as between the National Meteorological Services (NMS) and the National Hydrological Services (NHS) which tend to work many times in isolation. The general purpose of this work is to unravel the dysfunction/chaos of the way early warning systems are done and provide guidelines to integrate flood warning system at all scales to be used in operational forecasting, particularly for countries in the CA&CR. Flood warning can be seen as a set of sub-systems in which forecasting is only one of those sub-systems. A conceptual framework has been proposed to classify flood warning systems using the spatial and temporal scale at which the flood warning systems operate, subdividing them into Global, Regional, National and Local FEWS. In practice, these systems are not operated in an integrated manner. Emerging technology is available to allow the integration of global- and local-scale forecasting resources in the CA&CR. The Tethys Platform has a series of online tools applicable to flood forecasting. A workflow is given for the use of four apps in Tethys for flood forecasting: (i) Stream flow Prediction Tool; (ii) Reservoir operation tool; (iii) Hydro Viewer Hispaniola; Flood Map Visualization tool.

This thesis is focused in the representatives of beetle families Hydrophilidae and Hydraenidae of West Indies and adjacent regions. It consists of two parts, the phylogenetic part and the systematic part. The phylogenetic part focuses on the hydrophilid genera Phaenonotum and Crenitulus of Greater Antilles: beetles were sampled in all four main islands (Cuba, Hispaniola, Puerto Rico and Jamaica) and analyzed using the combination of molecular and morphological data. The genus Phaenonotum contains four single-island endemics, of which those from Cuba, Jamaica and Hispaniola are wingless and form a clade that diversified ca. 46 million years ago (Ma) and likely colonized the Caribbean via the GAARlandia land bridge. In contrast, the Puerto Rican endemic and the two remaining non-endemic species colonized the Greater Antilles by over- water dispersal during the Oligocene-Miocene. The analysis of the genus Crenitulus revealed that Greater Antillean species belong to two separate clades: the Crenitulus yunque clade endemic for Cuba and Hispaniola, and the Crenitulus suturalis clade containing specimens from Greater Antilles and from northern America. A detailed revision of the Crenitulus yunque clade using morphology and molecular-based species delimitation recognized 11 species locally endemic for...

Nickel silicate laterite deposits developed on ultra-mafic rocks are similar in many general respects but they vary considerably in detail. The mineralogy of these surficial deposits is very complex and difficult to determine because of the fine grained nature and solid solution characteristics of the hydrous secondary minerals and because many of the phases are actually mineraloids that are poorly ordered or amorphous. To try some new approaches toward clarification of these phases, 24 samples from New Caledonia and Puerto Rico ranging from the ophiolite-ultramafic olivine-pyroxene-chromite-serpentine substrate rocks upward through intermediate phases of weathering to the final oxide -hydroxide iron cap phase were analyzed with the infrared spectrophotometer (IR -10) and with the automated X –ray diffractometer. Four limonite samples were also mineralogically analyzed. Goethite, secondary quartz, cryptomelane, hematite, chromite, talc, thuringite, and garnierite have been identified in various samples as weathering profile products.