English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Disaggregation methods are designed to produce finer temporal scale data from coarser temporal scale data. An example of this need is the modification of monthly scale precipitation data to drive dynamic hydrological models operated at daily scale. Such issues occur for dynamic hydrological forecasts in which the General Circulation Models (GCMs) are used to generate the required precipitation data but the GCM outputs are only be reliable at coarser temporal scales. Stochastic weather generation methods together with simple adjustment of the total amounts provide a simple and efficient way for data disaggregation. However, the success of stochastic weather generation depends critically on whether the chain-dependent process can mimic the transient property between dry and wet days and whether the specified statistical distribution for amounts on wet days can capture the distributional properties of the amounts. Note that there are many small precipitation events within the catchment data which are treated differently for different purpose. Simple elimination by using different threshold values to define dry/wet days was frequently used in practice. A single statistical distribution may not be sufficient to capture the characteristics of amounts on positive data. Furthermore, the performance of dynamic hydrological model depends also on whether the specified statistical distribution for amounts can capture large precipitation events. Various proposals have been developed to better capture the precipitation pattern, including the use of high-order Markov Chain model for the occurrence pattern and the use of a mixed distribution for the amounts. In this paper, a hybrid stochastic weather generation method is proposed to simulate daily precipitation data based on the monthly data at catchment scale. Firstly, a multiple state Markov chain model is used for occurrence pattern as two state model may not reflect different weather patterns. Secondly, a truncated Gamma distribution is used for smaller precipitation amounts as usual and a censored extended Burr XII distribution is employed for larger amounts as it can better capture the extreme values. By doing this, one needs not deal with discontinuity in the distribution of precipitation amounts and ensure that the state and its corresponding amount are well matched. In order to take account of seasonality, the models are constructed for individual months. The proposed method is demonstrated by using catchment data from different climatic regions in Australia. Results show that (1) The proposed multiple state Markov chain model can capture the states reasonable well; (2) The seasonality is an important factor to be considered in the disaggregation; (3) The proposed distributions for different states can mimic the amounts well. As an regression based approach for the occurrences and distribution fitting approach for the amounts, the proposed method can be easily extended to include external predictors for generalization and/or improvements.

A stochastic weather generation method for temporal precipitation simulation