Comparison of Hydrology Component Spatial Correlation Model (Continuous Model) and Markov Chain (Case Study of Cikapundung Watershed, Saguling Reservoir and Cipanunjang Reservoir)

Water allocation in the watershed is a concern in developing
countries where there are limited water resources and greater
demands with more parties. Past data on water flow amount is
the only information available to estimate future water delivery.
One of the strategies used is to determine the exact number of
water discharge based on past recorded data. Future prediction
data is essential for sustainable management of water resources.
The objective of this research is to construct a hydrological
model to estimate or anticipate future water discharge using the
Discrete Chain Markov model where the results are then
compared with continuous discharge rain model by
implementing multiple linear regression method. Both
components (rain and discharge) are then modified into several
alternatives in 3 different research sites to analyze and compare
the combination of the results with the highest correlation
value. The inflow discharge data in Cipanunjang used in this
study is the data retrieved from 2000-2013; data of Cikapunung
Watershed are retrieved from 2003-2012. Meanwhile, data of
Saguling are extracted from 1986-2013. In spatial correlation
method, QQQQ combination was acquired in Cipanunjang,
while the best combination that can be acquired in
Cikapundung and Saguling watersheds is QQt-1PP. The
correlation values based on the spatial correlation method in
Cipanunjang, Cikapundung watershed, and Saguling were
0.882, 0.871 and 0.857, respectively. Meanwhile, the
correlation values of Discrete Markov method in the same
places were 0.899, 0.718 and 0.782, respectively. These results
suggest that this continuous method can be used to predict the
amount of water discharge in the future well enough and is
effectively used in Indonesia territories.