1. What is the importance of sub-daily climate data in climate impact analysis?
Sub-daily climate data, which captures variations in temperature, precipitation, and other weather variables at intervals of less than a day, is becoming increasingly important in climate impact analysis. This type of data provides a more detailed representation of local and regional climate conditions and temporal variations. It is crucial for evaluating the impacts of climate change on various sectors such as agriculture, water resources, energy production, and human health. Sub-daily data also helps in assessing adaptation strategies that depend on behavior or processes with high temporal dynamics, such as energy demand, labor activity, heat stress of crops, and flood events. Research has shown that using sub-daily climate data results in more robust and reliable impact assessments compared to using daily data. While most climate model data are available at daily resolution due to high storage requirements, the demand for sub-daily data is increasing due to lower costs for storage and computing resources. Various methods exist to disaggregate available daily climate data to sub-daily, including statistical methods, weather generators, and mechanistic approaches. The Teddy-Tool, a globally applicable tool for climate impact studies, uses statistical methods for temporal disaggregation of daily climate model data, preserving mass and energy of daily climate model data for each variable and considering regional and seasonal climate characteristics and physical consistency between variables. It has been used with bias corrected daily CMIP6 climate data for historical time periods and future scenarios from the ISIMIP, providing a framework for consistently projecting the impacts of climate change across affected sectors and spatial scales. The Teddy-Tool is available open source via Zenodo, making it accessible for climate impact assessments in different sectors.
read more
2. What is the procedure for disaggregating daily bias-corrected climate model data for air temperature, humidity, shortwave radiation, longwave radiation, air pressure, windspeed, and precipitation using Teddy?
The procedure for disaggregating daily bias-corrected climate model data for air temperature, humidity, shortwave radiation, longwave radiation, air pressure, windspeed, and precipitation using Teddy involves several steps. First, hourly WFDE5 data are aggregated to daily values and stored as NetCDF files. The daily aggregation uses mean values for all variables and daily sums for precipitation. Daily maximum and minimum temperature are calculated from the hourly data. Units of climate inputs are converted to match the Teddy output. For the conversion of specific humidity to relative humidity, the Buck equation is applied. The climate model day of interest and a basic population of 920 WFDE5 days are classified according to their precipitation state. Only days with the same precipitation class as the climate model day of interest are selected for further calculations. The absolute error between daily climate model and aggregated daily WFDE5 data for each variable is calculated, and the most similar meteorological day is determined based on the lowest cumulated ranks. The hourly values from the most similar day of the WFDE5 reference dataset are taken for each variable and divided by the WFDE5 daily mean value of the selected day, preserving the relative diurnal profiles. The daily mean is conserved, and temperature is further scaled between the provided minimum and maximum. Relative humidity is limited to 100%, and hourly values can be aggregated to the user-defined time step. In rare cases, precipitation cannot be distributed, and several options are implemented to handle exceptions. The calculation procedure can be performed either for universal time (UT) or for local solar time (LST).
read more
3. How does the DOY window size impact the correlation of precipitation in different regions?
The impact of the DOY window size on the correlation of precipitation varies between regions. Larger DOY windows are mainly beneficial for precipitation in tropical and arid regions, while in regions with pronounced seasons, the correlation might decrease with larger DOY window size. The results also show that the correlation for precipitation is generally larger in tropical regions than in continental regions. Additionally, the autocorrelation over lag times between one and 24 hours is calculated for precipitation and wind speed to statistically capture and validate sub-daily patterns and inter-hour connectivity. The number of wet hours and hours with low wind speed are also evaluated to assess the reproduction of these statistics with different DOY window sizes. Overall, the DOY window size plays a significant role in the correlation and reproduction of precipitation statistics in different regions.
read more
4. What are the advantages of using the Teddy-Tool for temporal disaggregation of daily climate model data?
The Teddy-Tool offers several advantages for temporal disaggregation of daily climate model data. Firstly, it does not require any additional input data, only the daily climate model data. Secondly, it considers specific regional and seasonal features of the diurnal course of different climate variables, ensuring high correlations (>0.9) for most variables, except for precipitation (>0.5) and wind speed (>0.75). The tool is relatively simple to apply and allows for the reproduction of mean daily values of the climate model at any time. It also conserves mass and energy, ensuring accurate results. The spatial and temporal resolution of the results is determined by the provided temporal and spatial resolution of the chosen reference data, such as WFDE5. Additionally, the Teddy-Tool can be used with other reference data that provides higher temporal or spatial resolution for a specific region. The tool's evaluation reveals that a DOY window size of 11 can generally be recommended across all variables, with larger window sizes being avoided in arid regions and shorter windows leading to poorer representations of autocorrelation and extreme events. However, the tool has limitations in representing extreme events, particularly for precipitation. For temporal disaggregation of extreme precipitation, dynamical downscaling via high-resolution climate models is recommended. Despite these limitations, the Teddy-Tool holds promise for temporal disaggregation under climate change, provided that the most similar day of historical data is selected and that the diurnal profile is representative of future climatic conditions. Future developments could include improved inter-day connectivity, temperature and surface pressure classes, and the separation of direct and diffuse radiation.
read more