Demand for freshwater is rising with factors, such as population growth, land use change and climate variations, rendering water availability in the future uncertain. Groundwater resources are being increasingly exploited to meet this growing demand. The aim of this study is to identify the influence of population growth induced by land use change and climate change on the future state of freshwater resources of Lamu Island in Kenya where a major port facility is under construction. The results of this study show that the “no industrial development” population scenario (assuming the port was not constructed) would be expected to reach ~50,000 people by 2050, while the projected population upon completion is expected to reach 1.25 million in the same year when the Lamu Port-South Sudan-Ethiopia Transport Corridor Program (LAPSSET) port reaches its full cargo-handling capacity. The groundwater abstraction in 2009 was 0.06 m³ daily per capita, while the demand is expected to raise to 0.1 m³ by 2050 according to the “LAPSSET development” projection. The modelling results show that the Shela aquifer in Lamu, which is the main source of water on the island, will not experience stress by 2065 for the “no industrial development” population scenario, whereas for the “LAPSSET development projection” population scenario, it will occur sooner (between 2020 and 2028). The modelling results show that the Representative Concentration Pathways (RCP) climate change scenarios will have a smaller impact on the effective water volume reserves than Special Report on Emissions Scenarios (SRES) for the “no industrial development”, while the impact is expected to be similar for the “LAPSSET development”, suggesting that population growth exacerbated by land use change will be a more significant driving force than climate change in affecting freshwater availability.

High-resolution electrical resistivity tomography (ERT) with electrode spacing of 1 m proved to be an effective methodology to characterize ephemeral, sparse, and discontinuous freshwater lenses within the almost completely salinized shallow aquifer of the low coastal zone near Ravenna. ERT profiles with a vertical resolution of 0.25–0.5 m, once calibrated with groundwater data (water table depth and electrical conductivity) collected in multi-level sampler systems at the same spatial resolution, are reliable and provide repeatable measurements in time. The ERT methodology allows for fast data acquisition over large areas and it also permits the study of the evolution in time of freshwater availability in coastal zones, which is important for local ecosystems and soil resources. This makes high-resolution ERT a valid tool to aid local stakeholders and decision makers to effectively manage freshwater lenses, and guarantee their preservation or augmentation by means of managed aquifer recharge.

Coastal dunes provide the hinterland with natural protection from marine dynamics. The specialized plant species that constitute dune vegetation communities are descriptive of the dune evolution status, which in turn reveals the ongoing coastal dynamics. The aims of this paper were to demonstrate the applicability of a low-cost unmanned aerial system for the classification of dune vegetation, in order to determine the level of detail achievable for the identification of vegetation communities and define the best-performing classification method for the dune environment according to pixel-based and object-based approaches. These goals were pursued by studying the north-Adriatic coastal dunes of Casal Borsetti (Ravenna, Italy). Four classification algorithms were applied to three-band orthoimages (red, green, and near-infrared). All classification maps were validated through ground truthing, and comparisons were performed for the three statistical methods, based on the k coefficient and on correctly and incorrectly classified pixel proportions of two maps. All classifications recognized the five vegetation classes considered, and high spatial resolution maps were produced (0.15 m). For both pixel-based and object-based methods, the support vector machine algorithm demonstrated a better accuracy for class recognition. The comparison revealed that an object approach is the better technique, although the required level of detail determines the final decision.