The Hindu-Kush Karakoram Himalaya (HKKH) region is called the "Third Pole" of our planet for the large amount of frozen water resources which are present in the area. Rivers from the HKKH and the Tibetan Plateau bring water to more than one billion people in South-East Asia; for this reason, the understanding of the ongoing and expected changes of the hydrological cycle in these mountain regions is the utmost importance. Precipitation is one of the key climatic variables for performing such a kind of investigations.
The pilot study described here is devoted to the use of both observational and climate model data to analyze the climatology of precipitation in HKKH, the long term precipitation trends, the foreseen future changes under different emission scenarios and, more specifically, the interaction between mid-latitude western disturbances and the tropical monsoonal circulation, the two main circulation patterns affecting the amount, spatial distribution and seasonality of precipitation in the HKKH region..
Fig 1: Multiannual mean (1998-2007) of winter (December to April) precipitation over the region between 69°-95°E and 23°-39°N from the APHRODITE, CRU, GPCC, TRMM, GPCP, ERA-Interim and the EC-Earth model datasets (Palazzi et al., 2013)
An analysis of precipitation in the HKKH region has been performed in a first study (Palazzi et al., 2013), using several datasets from different observational archives, the ERA-Interim reanalysis and the output of the global climate model (GCM) EC-Earth (the EC-Earth simulations have been performed at CNR-ISAC). The employed observational archives are based on in-situ gridded station data (APHRODITE, CRU, GPCC), on satellite data (TRMM), and on a merged in-situ/satellite climatology (GPCP); see for example Figure 1. Two specific sub-regions in the HKKH, which differ for circulation patterns, precipitation distribution and amounts and, by consequence, glacier behaviour and dynamics have been identified: the HKK (Hindu Kush-Karakoram) and the Himalaya regions. The analysis is focused on precipitation seasonality, interannual variability and long-term trends (the contribution of rain and snow is considered separately, when possible). The capability of the various datasets to reproduce precipitation characteristics has been evaluated and their biases and issues have been analyzed; the EC-Earth model has been validated against the observations and a process-oriented evaluation has been performed. The model has been used to analyse future precipitation trends in the HKK and Himalaya, based on the most recent IPCC emission scenarios “RCP 4.5” and “RCP 8.5”.
In a second study (Palazzi et al., 2014) we have extended the EC-Earth model analysis in the HKKH region to an ensemble of thirty-two GCMs participating in the World Climate Research Program (WCRP) Coupled Model Intercomparison Project Phase 5 (CMIP5), in order to provide an overview of their performance in simulating the current and future (out to 2100) climatology of precipitation in this area, to discuss the spread among the various models and highlight some of the factors responsible for differences in the models behaviour and performances. For validation purposes, we have compared the models outputs with the two longest precipitation datasets already considered in the first study going back in time to 1901. One is the last available Climate Research Unit (CRU) product, consisting of monthly gridded fields of precipitation from 1901 to 2012 over land areas with a spatial resolution of 0.5° latitude-longitude; the other is the Global Precipitation Climatology Centre (GPCC) full data reanalysis (GPCC_FD), providing data from 1901 to 2010 and consisting of monthly precipitation data at 0.5° spatial resolution. We have analysed the annual cycle climatology of precipitation and long-term historical trends in the two sub-regions of the HKKH domain, by evaluating and quantifying the spread of the models around their mean (multi-model mean). We have then analysed the future precipitation changes relative to present conditions, focusing on both near and far future decades, in the RCP4.5 and RCP8.5 emission scenarios. One of the outcomes of this study is that, in both regions, the model spread relative to the multi-model mean is large, indicating that the models do not converge in their representation of the historical precipitation annual cycle. Despite this, all models reproduce one-modal precipitation annual cycles in the Himalayan region, even if the various distributions are differently wide and have different amplitudes, while the model disagreement is much more serious in the HKK region, where annual cycles with very different characteristics. The disagreement among the models is particularly severe in summer, indicating a difficulty of the models in reproducing in a consistent way the effects of the summer monsoon circulation in the HKK sub-region (or in relatively small portion of it, located in the easternmost part). A hierarchical clustering analysis has been applied to group the various models based on their output in terms of precipitation annual cycle in the HKK region, so assuming no a priori knowledge about the features of any model, in order to identify what model features give rise to the various annual cycle climatologies reproduced by the models. An example of such a clustering procedure applied to the HKK region is shown in Figure 2.
Fig 2. Mean annual cycle of precipitation in the HKK simulated by all models within each cluster (the grey shaded areas indicate the variability range of the models) and by their multi-model mean (MMM, black lines with symbols). The CRU and GPCC observations are shown with the pink and green lines, respectively (Palazzi et al., 2014, under revsion. Please to not use the image)
Finally, in a third study (Filippi et al., 2014) we have analysed the mechanisms through which teleconnection patterns such as the North Atlantic Oscillation (NAO) influence winter precipitation in the HKK region. Previous studies, in fact, have indicated that winter precipitation and NAO are correlated with above (below) than normal precipitation over the HKK area during the positive (negative) NAO phase (see Figure 3). The main goal of our research is to identify the processes that are responsible for the observed correlation between NAO and winter precipitation in HKK and the secular changes occurred in the correlation in the past century.
Fig 3. Correlation coefficients between NAOI and winter precipitation from (a) GPCC, (b) CRU, (c) APHRODITE and (d) ERA40. Colours indicate statistically significant correlations at the 95% confidence level. Non-significant correlations are marked in grey. The green circle highlights the area of positive correlation centred on HKK (Filippi et al., 2014, minor revisions. Please do not use the image).
- Palazzi, E., von Hardenberg, J., & Provenzale, A. (2013). Precipitation in the hindu-kush karakoram himalaya: Observations and future scenarios. Journal of Geophysical Research D: Atmospheres, 118(1), 85-100
- Palazzi E., J. von Hardenberg, S. Terzago, A. Provenzale (2014). Precipitation in the Karakoram- Himalaya: A CMIP5 view. Accepted for publication, Climate Dynamics
- Filippi, L., E. Palazzi, J. von Hardenberg, A. Provenzale (2014). Multi-decadal variations in the relationship between NAO and winter precipitation in the Hindu-Kush Karakoram. Accepted for publication, Journal of Climate