Creation and conservation of urban parks is challenging in arid environments where daily thermal extremes, water scarcity, air pollution and shortage of natural green spaces are more conspicuous. Water scarcity in the arid regions of Iran is major challenge for water managers. Accurate estimation of urban landscape evapotranspiration is therefore critically important for cities located in naturally dry environments, to appropriately manage irrigation practices. This study investigated two factor-based approaches, Water Use Classifications of Landscape Species (WUCOLS) and Landscape Irrigation Management Program (LIMP), to measure the water demand of two heterogeneous urban landscapes: a botanic garden and a sparse forest park. The irrigation water volume applied was compared with the gross water demand for the period from 2011 to 2013. In this research, WUCOLS estimated the annual water requirement of a botanic garden and a sparse forest park to be 5% and 44% lower, respectively, than LIMP. Comparison of estimated and applied irrigation showed that water savings can be made by the LIMP method. The outcomes of this research stressed the need to modify the irrigation requirements based on effective rainfall throughout the year, rather relying on long-term average data.