A dynamic flare with anomalously dense flare loops

The dynamic flare of November 6, 1980 developed a rich system of growing loops which could be followed in H-alpha for 1.5 hours. Throughout the flare, these loops, near the limb, were seen in emission against the disk. Theoretical computations of b-values for a hydrogen atom reveal that this requires electron densities in the loops to be close to 10 to the 12th per cu cm. From measured widths of higher Balmer lines the density at the tops of the loops was found to be 4 x 10 to the 12th per cu cm if no nonthermal motions were present. It is now general knowledge that flare loops are initially observed in X-rays and become visible in H-alpha only after cooling. For such a high density a loop would cool through radiation from 10 to the 7th K to 10 to the 4th K within a few minutes so that the dense H-alpha loops should have heights very close to the heights of the X-ray loops. This, however, contradicts the observations obtained by the HXIS and FCS instruments on board SMM which show the X-ray loops at much higher altitudes than the loops in H-alpha. Therefore, the density must have been significantly smaller when the loops were formed and the flare loops were apparently both shrinking and becoming denser while cooling.