Rock weathering is a fundamental geological process which primes fresh rock to form soil by chemical, physical and biological processes. However, despite the importance of weathering, little is known about deep weathering and its dependence on climatic factors. In this thesis, deep weathering was investigated at four study sites in the Chilean Coastal Cordillera. For this purpose, drilling campaigns were performed to obtain undisturbed weathering profiles from fresh bedrock to soil. The sites are located in four climatically different settings (arid, semi-arid, mediterranean, humid), and thus permit investigation of the climate dependence of weathering depth and degree. The drill cores extend to a depth of ca. 90 m at the arid, semi-arid and mediterranean sites; at the humid site two ca. 50 m deep drill cores were obtained. In these five drill cores, deep weathering was analysed using geochemical methods, which include weathering parameters that indicate chemical mass loss. In addition, a new indicator for the retention of reactive elements in secondary weathering products is developed in this thesis using a sequential extraction procedure. Moreover, denudation rates (i.e. the rate of chemical and physical removal of material from a landscape) were determined with a method that has not yet been applied for this application: the isotope ratio (10Bemet/9Be) of meteoric beryllium-10 (10Bemet), a cosmogenic radionuclide, and the stable isotope beryllium-9 (9Be). The weathering indicators define weathering profiles that differ in their depth and degree along the climate gradient. Deep weathering was identified at the semi-arid and mediterranean sites. At the arid site, no weathering was observed due to the lack of precipitation. Rather, alterations caused by hydrothermal overprinting were present. This drill core was therefore taken as a reference for hydrothermal alterations that have not been overprinted by meteoric processes. At the humid site, the two drill cores show weathering only in the upper 15 m of the weathering profile. Primary mineral dissolution and the concentration of extracted reactive elements like aluminium and iron that are incorporated into secondary weathering products are highest at this site. The comparison between the four sites therefore indicates that deep weathering is caused by a variety of different factors. One main factor is the transport of water and gases (e.g. O2) to depth, which occurs via diffusion through pore spaces or advection through open fractures in the rock. Primary mineral dissolution creates secondary porosity and introduces new transport pathways for water and gases. Tectonic fractures connect the Earth's surface with the subsurface and enable fast advective transport of water to depth. At the semi-arid site, deep weathering was observed to a depth of ca. 77 m – mainly located along tectonic fractures – but a continuous weathering gradient was found only in the upper 10 m. Therefore, the advective transport of water and gases through fractures facilitates deep weathering at this site. Tectonic fractures also enable deep weathering at the mediterranean site (to at least 75 m), where a continuous weathering gradient was observed down to a depth of ca. 42 m. At the humid site, there are less open fractures and thus fewer direct transport pathways for water to depth are available. In addition, the high water availability results in intense precipitation of secondary weathering products that clog the porosity and prevent water flow to depth. The pre-conditioning of bedrock can further promote deep weathering. Different processes can prime the minerals in bedrock: hydrothermal overprinting, post-magmatic cooling processes, mineral dissolution due to deep groundwater flow, and reactions with deeply diffused oxygen. These processes facilitate the mobilisation of elements from primary minerals. With sufficient water flow through the weathering profile, these mobilised and soluble elements can be removed and lost into the dissolved phase. The applied sequential extraction of reactive elements from secondary weathering products is a promising new indicator for this potential pre-conditioning of primary minerals. Another factor is the time that is available for weathering processes. This time was determined by denudation rates derived from cosmogenic beryllium-10. Considering the time required for 10 m of the weathering profile to be removed, it takes ca. 900 000 years at the semi-arid location. Hence, deep weathering along fractures is possible even with the minute water flow at this site. At the mediterranean site, the denudation rate is significantly higher, such that the removal of 10 m takes ca. 200 000 years. Despite the shorter time available for weathering processes, a deep and continuous weathering gradient has formed due to the higher water flow in combination with tectonic fracturing, porosity formation and the pre-conditioning of primary minerals which was identified in this thesis by means of extractable elements. At the humid site, the removal of the upper 10 m of the weathering profile requires ca. 700 000 years. This time enables intensive near-surface weathering, as indicated by high primary mineral dissolution and precipitation of secondary phases, though the clogging of porosity prevents deep weathering. The 10Bemet/9Be isotope system was developed for calculating denudation rates from surface samples and applied as an indicator for the time that is available for weathering processes and the depth-dependence of weathering. In this thesis, it was found that below an annual precipitation of 400 mm the depositional flux of cosmogenic 10Bemet produced in the atmosphere cannot be determined with a global climate model, as these models overestimate precipitation and thus the deposition of the nuclide. Above this precipitation limit denudation rates determined from 10Bemet /9Be agree well with the more established method that uses in situ 10Be produced in quartz. Furthermore, the depth distribution of both isotopes shows that reactive 9Be can be used as an indicator of strong alteration, either by hydrothermal overprinting or intense primary mineral dissolution, even at depth. 10Bemet only infiltrates into the upper meters of a weathering profile and adsorbs onto reactive weathering products, which prevents infiltration to depth. This thesis presents new findings on deep weathering and tests new indicators to identify deep weathering. At two of four sites, the semi-arid and mediterranean, deep weathering was identified in the drill cores. Tectonic fractures facilitate the transport of water and gases to depth at both sites. High water availability and the resulting high vegetation density enable intensive weathering, especially in the uppermost meters below the surface. As fractures connect the surface with the subsurface, microorganisms as well as water and gases can potentially migrate to depth. Moreover, the chemical extractions indicate that mineral nutrients and inorganic reaction partners for microbial growth are also present at great depth when minerals are pre-conditioned.
