Articles | Volume 22, issue 4
Nonlin. Processes Geophys., 22, 383–402, 2015
Nonlin. Processes Geophys., 22, 383–402, 2015

Research article 16 Jul 2015

Research article | 16 Jul 2015

Spatial random downscaling of rainfall signals in Andean heterogeneous terrain

A. Posadas1,2, L. A. Duffaut Espinosa1,3, C. Yarlequé4, M. Carbajal1, H. Heidinger5, L. Carvalho5, C. Jones5, and R. Quiroz1 A. Posadas et al.
  • 1Production Systems and the Environment Division, International Potato Center (CIP), Lima, Peru
  • 2World Agroforestry Centre (ICRAF), Nairobi, Kenya
  • 3Electrical and Computer Engineering Department, George Mason University, Virginia, USA
  • 4Department of Atmospheric and Environmental Sciences, University at Albany, State University of New York, Albany, USA
  • 5Earth Research Institute (ERI), University of California Santa Barbara, Santa Barbara, USA

Abstract. Remotely sensed data are often used as proxies for indirect precipitation measures over data-scarce and complex-terrain areas such as the Peruvian Andes. Although this information might be appropriate for some research requirements, the extent at which local sites could be related to such information is very limited because of the resolution of the available satellite data. Downscaling techniques are used to bridge the gap between what climate modelers (global and regional) are able to provide and what decision-makers require (local). Precipitation downscaling improves the poor local representation of satellite data and helps end-users acquire more accurate estimates of water availability. Thus, a multifractal downscaling technique complemented by a heterogeneity filter was applied to TRMM (Tropical Rainfall Measuring Mission) 3B42 gridded data (spatial resolution ~ 28 km) from the Peruvian Andean high plateau or \textit{Altiplano} to generate downscaled rainfall fields that are relevant at an agricultural scale (spatial resolution ~ 1 km).