Climate Simulation in Eurasian Grasslands

 (Kazakhstan)

Archaeological context and methods:

As described, the goal of our research is to develop new methods and algorithms for modeling the relationship between climate variation and vegetation geography to produce a high resolution reconstruction of environmental change.  By correlating commonly cited factors such as precipitation, temperature, and elevation with variation in vegetative proxy values we are developing functions that accurately simulate landcover “index” values under prescribed climatic conditions in foothill and mountain grasslands of Central Asia.

Statistical analysis of the study zone’s NDVI values, Precipitation, Elevation, and Temperature reveals two key points.  First, Eurasian mountain grassland environments are not heavily affected by annual differences in average monthly temperature, at least not during their primary growth period (August). Second, greenness of grassland cohorts in Dzhungaria is well correlated to average changes in regional precipitation (pearson correlation .667) and with changes in elevation (pearson correlation .68). 

In comparison to the average ranges of NDVI recorded for particular grassland cohorts, calculating %ΔNDVI using our simulation algorithm illustrates the effect of botanical change given a controlled amount of variation in particular forces, such as rainfall.  Calculating the simulated percent change of grassland greenness (standardized according to elevation) under incremental changes in precipitation shows that lower, more arid grassland ranges (e.g. Artemisia/grassy steppe) are more highly responsive to nominal increases in precipitation, whereas highland meadow (fescue) ranges are little affected.

For example at 1000masl, a 2mm increase in average August precipitation correlates with a 45% increase in green vegetation, while the same increase at 2300masl will cause 5% increase.  This makes logical botanical sense, since arid grassland cohorts consist of a diversity of meso-xeric species that rapidly flourish with proportionally smaller increases in precipitation, while a small rainfall increase at higher elevations will have little impact on the growth and phenology of endemic grass cohorts in that zone. The advantage of the simulation algorithm is that all the variables that contribute to the productivity of various regional grassland cohorts can be modeled and iterated too isolate specific relationships between changing climatic conditions and landcover.

Discovery Highlights

Statistical analysis of the study zone’s NDVI values--Precipitation, Elevation, and Temperature--reveals two key points:

1. Eurasian mountain grassland environments are not heavily affected by annual differences in average monthly temperature, at least not during their primary growth period (August).

2. Greenness of grassland cohorts in Dzhungaria is well correlated to average changes in regional precipitation and with changes in elevation

In comparison to the average ranges of NDVI recorded for particular grassland cohorts, our simulation algorithm illustrates the effect of botanical change given a controlled amount of variation in particular forces, such as rainfall. 

Calculating the simulated percent change of grassland greenness (standardized according to elevation) under incremental changes in precipitation shows that lower, more arid grassland ranges are more highly responsive to nominal increases in precipitation, whereas highland meadow ranges are little affected. 

The advantage of the simulation algorithm is that all the variables that contribute to the productivity of various regional grassland cohorts can be modeled and iterated too isolate specific relationships between changing climatic conditions and landcover.