Examination of elevation dependency in observed and projected temperature change in the Upper Indus Basin and Western Himalaya

H. J. Fowler, N. D. Forsythe, S. Blenkinsop, D. Archer, A. Hardy, T. Janes, Richard Jones, T. Holderness

Research output: Contribution to conferenceAbstractpeer-review


We present results of two distinct, complementary analyses to assess evidence of elevation dependency in temperature change in the UIB (Karakoram, Eastern Hindu Kush) and wider WH. The first analysis component examines historical remotely-sensed land surface temperature (LST) from the second and third generation of the Advanced Very High Resolution Radiometer (AVHRR/2, AVHRR/3) instrument flown on NOAA satellite platforms since the mid-1980s through present day. The high spatial resolution (<4km) from AVHRR instrument enables precise consideration of the relationship between estimated LST and surface topography. The LST data product was developed as part of initiative to produce continuous time-series for key remotely sensed spatial products (LST, snow covered area, cloud cover, NDVI) extending as far back into the historical record as feasible. Context for the AVHRR LST data product is provided by results of bias assessment and validation procedures against both available local observations, both manned and automatic weather stations. Local observations provide meaningful validation and bias assessment of the vertical gradients found in the AVHRR LST as the elevation range from the lowest manned meteorological station (at 1460m asl) to the highest automatic weather station (4733m asl) covers much of the key range yielding runoff from seasonal snowmelt. Furthermore the common available record period of these stations (1995 to 2007) enables assessment not only of the AVHRR LST but also performance comparisons with the more recent MODIS LST data product. A range of spatial aggregations (from minor tributary catchments to primary basin headwaters) is performed to assess regional homogeneity and identify potential latitudinal or longitudinal gradients in elevation dependency. The second analysis component investigates elevation dependency, including its uncertainty, in projected temperature change trajectories in the downscaling of a seventeen member Global Climate Model (GCM) perturbed physics ensemble (PPE) of transient (130-year) simulations using a moderate resolution (25km) regional climate model (RCM). The GCM ensemble is the17-member QUMP (Quantifying Uncertainty in Model Projections) ensemble and the downscaling is done using HadRM3P, part of the PRECIS regional climate modelling system. Both the RCM and GCMs are models developed the UK Met Office Hadley Centre and are based on the HadCM3 GCM. Use of the multi-member PPE enables quantification of uncertainty in projected temperature change while the spatial resolution of RCM improves insight into the role of elevation in projected rates of change. Furthermore comparison with the results of the remote sensing analysis component - considered to provide an 'observed climatology' - permits evaluation of individual ensemble members with regards to biases in spatial gradients in temperature as well timing and magnitude of annual cycles.
Original languageEnglish
Pagesabstract #A11A-0011
Publication statusPublished - 01 Dec 2013
EventAmerican Geophysical Union, Fall Meeting 2013 - San Francisco, California, United States of America
Duration: 09 Dec 201313 Dec 2013


ConferenceAmerican Geophysical Union, Fall Meeting 2013
Country/TerritoryUnited States of America
Period09 Dec 201313 Dec 2013


  • 3355 ATMOSPHERIC PROCESSES Regional modeling
  • 1637 GLOBAL CHANGE Regional climate change
  • 1640 GLOBAL CHANGE Remote sensing


Dive into the research topics of 'Examination of elevation dependency in observed and projected temperature change in the Upper Indus Basin and Western Himalaya'. Together they form a unique fingerprint.

Cite this