Weather or not—Global climate databases: Reliable on tropical mountains?

another global climate database that focuses on providing such high-resolution climate data. It uses a combination of downscaled reanalysis data and satellite data to produce spatially explicit climate information. Many studies published in high-impact journals, for example, have used WC data for their analyses. These studies often address topics such as species distribution, climate change impacts, ecosystem responses and environmental management strategies. In PLOS ONE, WC has been used in 920 studies since 2007, in PNAS in 131 studies since 2009, in Science in 49 studies since 2010, and in Nature in 44 studies since 2015. However, many tropical regions lack consistent observations of even the most basic meteo- rological parameters such as temperature, humidity or precipitation. This holds particularly true for areas of complex terrain, such as the region of Kilimanjaro, Tanzania. On the other hand, it is exactly these regions that are of high ecological interest and value with regard to bio- diversity and related ecosystem services, often termed ‘global biodiversity hot-spots’, where the accuracy of such data is an indispensable requirement for any meaningful analysis. Relying on results and conclusions based on above mentioned climate data sets modelled on a weak basis could be risky. Therefore, the quality of these tools needs to be assessed. However, proving the reliability of such data in remote tropical areas is mostly not possible due to the lack of stations, which poses a kind of dilemma.

In this study, we aim to verify the accuracy of WC and C–the two most frequently used data sets in biological research, with the highest available resolution—at 16 peaks on nine mountain ranges in Tanzania, with focus on Kilimanjaro, the highest mountain in Africa and the highest solitary mountain in the world using a unique climate dataset collected over the last 30 years mainly in the ecoclimatically most important montane and alpine zones. Weather stations, mainly for measuring precipitation, were already established in Tanzania in colonial times [11,12], but almost exclusively in the populated areas below the forest belts. Within the framework of several ecological projects of the German Research Foundation DFG, we were able to establish a network of weather stations on Kilimanjaro, Meru, the North Pare, the South Pare, West Usambara, East Usambara and Nguru (Figs 1 and 2) from 1996 onwards. The extensive climate network on Kilimanjaro, which extends from 700 to 4600m (Fig 3), was later partially integrated into the DFG research units KiLi (“Kilimanjaro under global change”) and later Kili-SES (“The role of nature for human well-being in the Kilimanjaro Social-Ecologi- cal System”), the latter project still ongoing. It includes 75 rainfall stations (Fig 3, S1 Table), where we mainly measure temperature, humidity, partly also wind speed and fog and in addi- tion 33 stations, where we only measure temperature and humidity. For this study we have also included available data from the Tanzania Meteorological Agency, private companies and

church missions from elevations below 1700 m (Fig 3, S1 Table), thus extending our own 75 rainfall stations to 125. We established the 21 stations on the other mountains after 2010 and complemented them with 26 stations of other sources (S1 Table), ranging from 120 to 2450m. Precipitation is one of the most fundamental meteorological elements and a highly dynamic one that changes its intensity, frequency and duration under the influence of topographic parameters such as elevation, vegetation and land use and is therefore more difficult to predict than temperature [13]. Therefore, and since the available dataset on precipitation is more exten- sive than that on other parameters, we here focus on mean annual precipitation (MAP). Materials and methods Study sites The mountains studied lie on a transect that extends 350 km inland from the Tanzanian coast. Fourteen of the sixteen peaks studied belong to the Eastern Arc, a chain of ancient crystalline PLOS ONE Global climate datasets on tropical mountains PLOS ONE | https://doi.org/10.1371/journal.pone.0299363 March 13, 2024 2 / 12 2962610 Email Address: met@meteo.go.tz Website: http://www.meteo.go.tz/. Funding: AH received funds from Deutsche Forschungsgemeinschaft (DFG) (https://www.dfg. de/), grant number HE 2719/14-1. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing interests: The authors have declared that no competing interests exist.