Characterization and modeling of hydrology and climatology in paramo ecosystems in Colombia : A study of the paramos and climate change impact on these ecosystems as a water resource

Abstract: Paramos are unique ecosystems situated in the tropical Andes at an elevation of 2800-4800 meters above sea level. They are home to a variety of endemic species and have a great water regulating capacity and supply around 80 % of the population in Colombia with freshwater. These ecosystems have been found to be particularly vulnerable towards climate change, which is a major challenge when it comes to safeguarding this water source. This study aimed to investigate paramos on three spatial scales: the large, medium and small scale. The large scale aimed at investigating the paramos of Colombia on a national level, the medium scale aimed at modeling the hydrology of a small paramo catchment, and the small scale aimed at exploring the general landscape and methods for monitoring the hydroclimate. At the national scale, the paramos were characterized based on hydrological and climatological variables, in the end focusing on precipitation and discharge, and producing descriptive statistics, in addition to finding temporal trends. The aim was to describe the paramos of Colombia hydroclimatologically on a spatial and temporal scale to get a more updated and comprehensive picture of the hydrology in these ecosystems and to investigate whether any possible relationships between the hydroclimatic variables could be detected that could affect the water availability, that is, if there were any signs of climatic forcing and hydrological response. Another aim was to investigate whether any significant trends could be seen in historical data and if there were any signs of climate change or ENSO (El Niño Southern Oscillation) events affecting the time series. The results showed that there was a large range in the amount of annual precipitation, between 500 and 5000 mm/yr, depending on the location of the precipitation station. The annual precipitation differed with up to around 2000 and 3500 mm/yr within and between different paramo regions, respectively. The annual mean daily discharge ranged between 0 and 25-30 m3/s and the total annual specific discharge ranged between 0 and 2500 mm/yr. It was not possible to detect any general, common temporal trend in neither precipitation nor discharge for all of the stations and the majority of the trends for individual stations were not significant. However, most locations that did present significant trends in precipitation showed an increase rather than a decrease, on a monthly rather than an annual scale. The trends that were significant for discharge were also increasing. No evident relationship between precipitation and discharge could be detected in neither amount nor pattern or trend over time, although some similarities and common traits could be found. Some local maxima and minima in precipitation and discharge coincided with ENSO events, but whether the significant trends are due to climate change remains unclear. The changes that were seen could have an impact on societies that are close to and rely on the paramos, but more studies on a seasonal and monthly scale are needed. At the medium scale, the HBV light model was used to simulate the hydrology in the small paramo catchment called Monterredondo. The aim was to investigate how a parsimonious model like the HBV model would perform compared to the more comprehensive SWAT model, which previously succeeded in modeling the hydrology in this watershed. Since the data is scarce in these regions, a simpler model could be more desirable. The results showed that the HBV was sufficient for the task, receiving a NSE value of 0.81 (calibration period) and 0.56 (validation period), and able to perform almost as well as the SWAT model, which received NSE values of 0.83 (calibration period) and 0.59 (validation period). Furthermore, since the paramos are vulnerable towards climate change, it was considered important to investigate how discharge might be affected by climate change. Simulations of future climate towards the end of the century, representing two emission scenarios, SSP2-4.5 and SSP5-8.5, were performed with data from the general circulation models EC-Earth3-CC (Europe) and MRI-ESM2-0 (Japan). The results from the EC-Earth3-CC model indicate a change towards drier dry seasons and wetter wet seasons. On an annual scale, it shows a decrease in discharge, but no change in annual precipitation. The MRI-ESM2-0 model was not considered reliable due to the large distance between the catchment area in question and the location where the data came from. Large uncertainties are associated with the data quality and quantity for both parts of this study. This is a well-known issue in these quite unapproachable locations, which is why the small scale investigation focused on learning more about the monitoring of these ecosystems. A horizontal precipitation station was built to capture horizontal precipitation. Even if built as a rather rough structure, it seemed to work and could be useful for future investigations. The results showed that horizontal precipitation contributed with a non-negligible amount to the total precipitation.  In summary, there is a risk that climate changes will have an impact on the water resources, due to drier dry seasons and wetter wet seasons. This could lead to water shortages in periods, but more studies are necessary to confirm the result. It is especially important to investigate more variables such as land use, soil moisture and radiation. It is also important to study more paramo areas in detail since they respond differently towards climate change.