干旱区生态环境知识资源中心

We have explored the potential of using magnetic susceptibility (chi(If)) as a proxy for past rainfall variations in a tropical region, and also the potential of tank-bed sediments from Southern India as a paleoarchive. The 3,700-year. rock magnetic record from Thimmannanayakanakere (14 degrees 12'N; 76 degrees 24'E), a small tank near Chitradurga town, Karnataka, exhibits consider able fluctuations in chi(If). Discounting the possibilities of magnetite from biogenic and anthropogenic sources and of dissolution of magnetic minerals, the chi(If) signal must be related principally to rainfall in the tank-catchment. There is a good correlation of chi(If) with instrumental rainfall data (r = 0.65 for Peninsular India; r = 0.45 for Chitradurga Station) and historically recorded rainfall events. Presuming that chi(If) variations are produced mainly by rainfall variations and that there was no significant human impact in the catchment and no slumping or bioturbation of sediments, we have reconstructed the paleorainfall profile for the last 3,700 years for Chitradurga region. The profile is chronologically constrained by two C-14 dates. Several events of drought and high rainfall, and onset of aridity are correlatable with similar events documented in speleothems from different parts of India, a tree-ring from western India and lake sediments from Rajasthan. We have thus shown that chi(If) is a proxy for rainfall in tropical regions and that tank-bed sediments from tropical India are an important archive of paleorainfall/paleoclimate data.