Forests immersed in high frequency of fog are commonly known as cloud forest. The presence of fog in cloud forests not only affects the water cycle, but heavily impacts forest structure. Cloud forests serve as vital habitats for many endemic species, including epiphytes, orchids, birds, and mammals. The high biodiversity characteristic of cloud forest makes it become the unique and important ecosystem (Aldrich et al. 1997;
Bruijnzeel et al. 2011). For example, the richness of endemic species in cloud forests was four times more than non-cloud forests of similar geographic and topographic settings (Bruijnzeel et al. 2010). Besides, fog intercepted by trunks and foliar surfaces provide not only the ecosystem additional water resources apart from rainfall but bring essential fertilizing ions to the plants rooting in the canopy and on the ground (Foster 2001;
Gottlieb et al. 2019). In the cloud forest of high altitudes, the fog water intercepted by plants may also play a pivotal role in supplying water of downstream regions (Stadtmüller 1987). Recent reports depicted that montane cloud forests were threatened due to the prevailing trend of elevated temperatures (Helmer et al. 2019). Evidences suggested that the rising of sea surface temperature is associated with fog belt shifting upward, which not only declines montane cloud forest areas but risks survival of endemic species (Ponce-Reyes et al. 2012; Pounds et al. 1999; Still et al. 1999). In addition, the general circulation models, which is popular with projecting future climatic parameters trend, suggested the scenario of warming climate in the coming decades in tropical mountains (Oliveira et al.
2014). In fact, temperature changes may cause stress on cloud forests such as shifting cloud base, altering evapotranspiration rate, impacting on hydrological cycle, and changing dominant species. In sum, cloud forests, especially tropical montane cloud
forest, which include invaluable ecological resources, are highly vulnerable to climate change (Ponce-Reyes et al. 2012).
Fog is officially defined as the visibility less than one kilometer when cloud touch the surface (NOAA 2005). In practice, the definition of fog vary from case to case (Izett et al.
2018). Its various definitions partly result from its hard-to-quantify characteristic; the first global cloud forest map was published by Bubb et al. (2004) using satellite remote sensing.
In the first cloud forest map, Asia (Insular Asia-Pacific, India, Sri Lanka, and Australia) takes the highest area proportion (nearly 60%) comparing to other continents. However, the hot spots of cloud forest studies regarding to hydrology and biodiversity are mainly conducted at Mesoamerican such as Monteverde, Costa Rica and Barro Colorado Island, Panama (Bruijnzeel 2001) of Central America. As the dynamics of fog largely being controlled by local environmental factors, the scenarios vary geographically. For example, a small mountain will experience steeper temperature lapse rate, lower sea surface temperature, and higher relative humidity than large mountain (Foster 2010). Thus, it is challenging to apply a single environmental condition to study the fog dynamic in one place, and it is urgent and needed to study fog dynamic especially in Asia.
So far, the worldwide cloud forest maps are restricted their estimated regions in the tropical areas (Bubb et al. 2004; Mulligan and Burke 2005; Wilson and Jetz 2016). It seems rational because tropical regions have the highest proportional cloud forest distribution comparing to higher latitudes areas. However, this restriction may also ignore places where are cloud forest ecosystems but outside tropics. Beyond that, it is common
cloud forest ecosystems. In fact, montane cloud forest in Taiwan not only occur in the tropics but be observed around the entire island with most of them outside the defined tropic (Schulz et al. 2017).
To date, montane cloud forest distribution can be mapped by conflating data from satellite, vegetation species, and meteorological stations with coarse resolution (more than one km).
These results reveal potential location of cloud forest globally, but lack the detailed spatial dynamic of fog at local scale. Thus, more fine-scale studies are needed to understand fog dynamics in heterogeneous mountainous regions. After all, comprehending fog dynamics reveals more detail about fog-forest interaction and their ecological effect (Weathers et al.
1986) than only identifying their location.
1.1. Research objectives
Chilan Mountain, which serves as a typical montane cloud forest ecosystem located in northeastern Taiwan, Pacific Asia, was selected as the study region. The orographic fog is more common in Taiwan than other fog types such as radiation fog in many other countries (Schulz et al. 2017). The diurnal cycle in our study area is that the air masses, which mostly come from the southeast direction lifted by orography in the evening, are cooled and condensed at the middle elevations (Klemm et al. 2006).
This study accesses this prevail fog formation (defined as cloud that touch the ground) to conduct fog duration research and attempts to achieve two objectives:
(1) Integrating satellite and in-situ meteorological data to develop the Fog Event Index (FEI), which have the ability to detect high spatiotemporal fog events.
(2) Investigating the spatiotemporal dynamics of fog duration in Chilan Mountain.
We designed a new approach, which by means of solar intensity difference as well as dew point depression to estimated 10 minutes resolution fog duration in a two year range across ~15.5 km wide region. We constrained our analyzed time within day time because the photosynthesis only happens during the day. The results can then reflect how fog interferes forest production potentially. To our knowledge, no any research have estimated this high spatiotemporal and long term period of fog dynamic at our sites before. Besides, our approach cannot provide the exact fog quantity but the highly temporal resolution duration. With the high temporal resolution fog duration along the altitudes, we can know how fog vary in montane areas, and provide a basic understanding of how much water and nutrients it potentially bring. Our estimated approach is purposed to reveal the mystery of fog in the montane cloud forest.