| Author | Kulatunga, Piyadasa |
| Call Number | AIT Thesis no. WA-83-23 |
| Subject(s) | Water resources development--Mathematical models
|
| Note | A thesis submitted in partial fulfillment of the requirement for the degree of Master of Engineering, School of Engineering and Technology |
| Publisher | Asian Institute of Technology |
| Abstract | There are many stochastic models for synthesizing stream flows for operational applications in water resources developments. The disaggregation model has become the latest model capable of maintaining the desired statistical resemblance between historical and synthetic seasonal (monthly) flows. Comparison of two approaches of preserving skewness in the disaggregation model , the Todini approach and the Hoshi and Burges (HB) approach are
given using monthly flows of Kirindi Oya at Lunugamvehera in Sri Lanka , recorded between 1944 andl979.
The first- order autoregressive model , known as an AR(l) process, which is capable of reproducing the observed short- term behavior of annual stream
flows is used to generate annual flows in the transformed domain using a three parameter log normal distribution. These annual flows are disaggregated
using the above mentioned approaches. The Sequent Peak algorithm is used to determine the necessary reservoir size to meet a given demand schedule, and storage probability curves are used to compare the two approaches.
The generated statistics from both approaches for 500 traces. compared with historical statistics. Both the approaches yielded slightly higher values of
mean for monthly flows while variances became smaller. This is due to the truncation of generated negative flows to zero. The variances and skewness coefficients of Todini ' s approach yielded smaller values than those of HB
approach. The skewness coefficients of generated flows in HB approach have good agreement with historical values. Both approaches preserved the within-year correlations between successive months and the correlation structure between monthly a nd annual flows.
Todini ' s approach has given smaller storages for all the three levels of annual demand examined . Incapability of preserving the historical variances in Todini ' s disaggregation scheme has more influence on reservoir size determination than the skewness coefficients of monthly flows. Todini's approach would give rise to spurious results of storage-reliability relationships, because the
variances and skewness coefficients of generated monthly flows become much
smaller than those of historical flows. Hence the HB approach of preserving skewness in disaggregation schemes is found to have a reasonable basis. |
| Year | 1983 |
| Type | Thesis |
| School | School of Engineering and Technology |
| Department | Department of Civil and Infrastucture Engineering (DCIE) |
| Academic Program/FoS | Water Resources Research Engineering (WA) |
| Chairperson(s) | Hoshi, Kiyoshi
|
| Examination Committee(s) | Huynh, Ngoc Phien ; Gupta, Ashim Das |
| Scholarship Donor(s) | CDG, The Federal Republic of Germany |
| Degree | Thesis (M.Eng.) - Asian Institute of Technology, 1983 |